2019-11-25 16:41:44 +01:00
-------------------------------------------------------------------
Mon Nov 25 14:59:24 UTC 2019 - Joao Cavalheiro <jcavalheiro@suse.com>
- Patch macros on spec file to support builds on SLE 12
2019-11-25 14:10:48 +01:00
-------------------------------------------------------------------
Thu Nov 21 13:29:22 UTC 2019 - Jan Fajerski <jan.fajerski@suse.com>
- Remove prometheus.firewall.xml source file
2019-11-21 14:26:06 +01:00
-------------------------------------------------------------------
Wed Nov 20 22:33:20 UTC 2019 - Michał Rostecki <mrostecki@opensuse.org>
- Remove firewalld files. They are installed in the main firewalld
package.
2019-11-21 14:23:19 +01:00
-------------------------------------------------------------------
Wed Nov 20 15:32:20 UTC 2019 - Joao Cavalheiro <jcavalheiro@suse.com>
- Update Uyuni/SUSE Manager service discovery patch
+ Modified 0003-Add-Uyuni-service-discovery.patch
+ Fixes crashes when systems have no FQDN
+ Adds Parallel calls to Uyuni API, meaningful performance increase
+ Adds Support for system group labels
2019-09-25 08:49:15 +02:00
-------------------------------------------------------------------
Mon Sep 23 10:19:03 UTC 2019 - Michał Rostecki <mrostecki@opensuse.org>
- Do not install the firewalld config file on Tumbleweed (on
versions newer than Leap 15.1). It's installed in the main
firewalld package.
2019-08-16 10:40:50 +02:00
-------------------------------------------------------------------
Fri Aug 16 06:46:24 UTC 2019 - Jan Fajerski <jan.fajerski@suse.com>
- remove 0004-prometheus-buildmode-pie.patch
+ this patch produces binaries that crash immediately
- add -v to promu build command for better build debugging
- reorder some %install tasks
Accepting request 720641 from home:jfajerski:branches:server:monitoring
- Add network-online (Wants and After) dependency to systemd unit bsc#1143913
- Build with PIE
+ Added 0004-prometheus-buildmode-pie.patch
- Only package required files (reduces rpm size by 4 MB)
- Add sysconfig file
- Add firewall config file
- Use variables for defining user and group
- Add support for Uyuni/SUSE Manager service discovery
+ Added 0003-Add-Uyuni-service-discovery.patch
- readded _service file removed in error.
- Update to 2.11.1
+ Bug Fix:
* Fix potential panic when prometheus is watching multiple zookeeper paths.
- Update to 2.11.0
+ Bug Fix:
* resolve race condition in maxGauge.
* Fix ZooKeeper connection leak.
* Improved atomicity of .tmp block replacement during compaction for usual case.
* Fix "unknown series references" after clean shutdown.
* Re-calculate block size when calling block.Delete.
* Fix unsafe snapshots with head block.
* prometheus_tsdb_compactions_failed_total is now incremented on any compaction failure.
+ Changes:
* Remove max_retries from queue_config (it has been unused since rewriting remote-write to utilize the write-ahead-log)
* The meta file BlockStats no longer holds size information. This is now dynamically calculated and kept in memory. It also includes the meta file size which was not included before
* Renamed metric from prometheus_tsdb_wal_reader_corruption_errors to prometheus_tsdb_wal_reader_corruption_errors_total
+ Features:
* Add option to use Alertmanager API v2.
* Added humanizePercentage function for templates.
* Include InitContainers in Kubernetes Service Discovery.
* Provide option to compress WAL records using Snappy.
+ Enhancements:
* Create new clean segment when starting the WAL.
* Reduce allocations in PromQL aggregations.
* Add storage warnings to LabelValues and LabelNames API results.
* Add prometheus_http_requests_total metric.
* Enable openbsd/arm build.
* Remote-write allocation improvements.
* Query performance improvement: Efficient iteration and search in HashForLabels and HashWithoutLabels.
* Allow injection of arbitrary headers in promtool.
* Allow passing external_labels in alert unit tests groups.
* Allows globs for rules when unit testing.
* Improved postings intersection matching.
* Reduced disk usage for WAL for small setups.
* Optimize queries using regexp for set lookups.
- rebase patch002-Default-settings.patch
- Update to 2.10.0:
+ Bug Fixes:
* TSDB: Don't panic when running out of disk space and recover nicely from the condition
* TSDB: Correctly handle empty labels.
* TSDB: Don't crash on an unknown tombstone reference.
* Storage/remote: Remove queue-manager specific metrics if queue no longer exists.
* PromQL: Correctly display {__name__="a"}.
* Discovery/kubernetes: Use service rather than ingress as the name for the service workqueue.
* Discovery/azure: Don't panic on a VM with a public IP.
* Web: Fixed Content-Type for js and css instead of using /etc/mime.types.
* API: Encode alert values as string to correctly represent Inf/NaN.
+ Features:
* Template expansion: Make external labels available as $externalLabels in alert and console template expansion.
* TSDB: Add prometheus_tsdb_wal_segment_current metric for the WAL segment index that TSDB is currently writing to. tsdb
* Scrape: Add scrape_series_added per-scrape metric. #5546
+ Enhancements
* Discovery/kubernetes: Add labels __meta_kubernetes_endpoint_node_name and __meta_kubernetes_endpoint_hostname.
* Discovery/azure: Add label __meta_azure_machine_public_ip.
* TSDB: Simplify mergedPostings.Seek, resulting in better performance if there are many posting lists. tsdb
* Log filesystem type on startup.
* Cmd/promtool: Use POST requests for Query and QueryRange. client_golang
* Web: Sort alerts by group name.
* Console templates: Add convenience variables $rawParams, $params, $path.
- Upadte to 2.9.2
+ Bug Fixes:
* Make sure subquery range is taken into account for selection
* Exhaust every request body before closing it
* Cmd/promtool: return errors from rule evaluations
* Remote Storage: string interner should not panic in release
* Fix memory allocation regression in mergedPostings.Seek tsdb
- Update to 2.9.1
+ Bug Fixes:
* Discovery/kubernetes: fix missing label sanitization
* Remote_write: Prevent reshard concurrent with calling stop
- Update to 2.9.0
+ Feature:
* Add honor_timestamps scrape option.
+ Enhancements:
* Update Consul to support catalog.ServiceMultipleTags.
* Discovery/kubernetes: add present labels for labels/annotations.
* OpenStack SD: Add ProjectID and UserID meta labels.
* Add GODEBUG and retention to the runtime page.
* Add support for POSTing to /series endpoint.
* Support PUT methods for Lifecycle and Admin APIs.
* Scrape: Add global jitter for HA server.
* Check for cancellation on every step of a range evaluation.
* String interning for labels & values in the remote_write path.
* Don't lose the scrape cache on a failed scrape.
* Reload cert files from disk automatically. common
* Use fixed length millisecond timestamp format for logs. common
* Performance improvements for postings.
Bug Fixes:
* Remote Write: fix checkpoint reading.
* Check if label value is valid when unmarshaling external labels from YAML.
* Promparse: sort all labels when parsing.
* Reload rules: copy state on both name and labels.
* Exponentation operator to drop metric name in result of operation.
* Config: resolve more file paths.
* Promtool: resolve relative paths in alert test files.
* Set TLSHandshakeTimeout in HTTP transport. common
* Use fsync to be more resilient to machine crashes.
* Keep series that are still in WAL in checkpoints.
- Update to 2.8.1
+ Bug Fixes
* Display the job labels in /targets which was removed accidentally
- Update to 2.8.0
+ Change:
* This release uses Write-Ahead Logging (WAL) for the remote_write API. This currently causes a slight increase in memory usage, which will be addressed in future releases.
* Default time retention is used only when no size based retention is specified. These are flags where time retention is specified by the flag --storage.tsdb.retention and size retention by --storage.tsdb.retention.size.
* prometheus_tsdb_storage_blocks_bytes_total is now prometheus_tsdb_storage_blocks_bytes.
+ Feature:
* (EXPERIMENTAL) Time overlapping blocks are now allowed; vertical compaction and vertical query merge. It is an optional feature which is controlled by the --storage.tsdb.allow-overlapping-blocks flag, disabled by default.
+ Enhancements:
* Use the WAL for remote_write API.
* Query performance improvements.
* UI enhancements with upgrade to Bootstrap 4.
* Reduce time that Alertmanagers are in flux when reloaded.
* Limit number of metrics displayed on UI to 10000.
* (1) Remember All/Unhealthy choice on target-overview when reloading page. (2) Resize text-input area on Graph page on mouseclick.
* In histogram_quantile merge buckets with equivalent le values.
* Show list of offending labels in the error message in many-to-many scenarios.
* Show Storage Retention criteria in effect on /status page.
+ Bug Fixes:
+ Fix sorting of rule groups.
+ Fix support for password_file and bearer_token_file in Kubernetes SD.
+ Scrape: catch errors when creating HTTP clients
+ Adds new metrics:
prometheus_target_scrape_pools_total
prometheus_target_scrape_pools_failed_total
prometheus_target_scrape_pool_reloads_total
prometheus_target_scrape_pool_reloads_failed_total
+ Fix panic when aggregator param is not a literal.
- fix spec file: actually ship promtool
- Update to 2.7.1:
+ Bug Fixes:
* Fix a Stored DOM XSS vulnerability with query history (boo#1124610)
* prometheus_rule_group_last_duration_seconds now reports seconds instead of nanoseconds
* Make sure the targets are consistently sorted in the targets page
- Update to 2.7.0:
+ cli flag depreacted: storage.tsdb.retention use storage.tsdb.retention.time
instead; depreacted flag will be removed in 3.0
+ Features:
* Add subqueries to PromQL
* Add support for disk size based retention. Note that we don't
consider the WAL size which could be significant and the time
based retention policy also applies (experimental)
* Add CORS origin flag
+ Bug Fixes:
* Don't depend on given order when comparing samples in alert unit testing
* Make sure the retention period doesn't overflow
* Don't generate blocks with no samples
- Update to 2.6.0:
+ Remove default flags from the container's entrypoint, run Prometheus from
/etc/prometheus and symlink the storage directory to /etc/prometheus/data
+ Promtool: Remove the update command
+ Features:
* Add JSON log format via the --log.format flag
* API: Add /api/v1/labels endpoint to get all label names
* Web: Allow setting the page's title via the --web.ui-title flag
+ Enhancements:
* Add prometheus_tsdb_lowest_timestamp_seconds, prometheus_tsdb_head_min_time_seconds
and prometheus_tsdb_head_max_time_seconds metrics
* Add rule_group_last_evaluation_timestamp_seconds metric
* Add prometheus_template_text_expansion_failures_total and prometheus_template_text_expansions_total metrics
* Set consistent User-Agent header in outgoing requests
* Azure SD: Error out at load time when authentication parameters are missing
* EC2 SD: Add the machine's private DNS name to the discovery metadata
* EC2 SD: Add the operating system's platform to the discovery metadata
* Kubernetes SD: Add the pod's phase to the discovery metadata
* Kubernetes SD: Log Kubernetes messages
* Promtool: Collect CPU and trace profiles
* Promtool: Support writing output as JSON
* Remote Read: Return available data if remote read fails partially
* Remote Write: Improve queue performance
* Remote Write: Add min_shards parameter to set the minimum number of shards
* TSDB: Improve WAL reading
* TSDB: Memory improvements
* Web: Log stack traces on panic
* Web UI: Add copy to clipboard button for configuration
* Web UI: Support console queries at specific times
* Web UI: group targets by job then instance
+ Bug Fixes:
* Deduplicate handler labels for HTTP metrics
* Fix leaked queriers causing shutdowns to hang
* Fix configuration loading panics on nil pointer slice elements
* API: Correctly skip mismatching targets on /api/v1/targets/metadata
* API: Better rounding for incoming query timestamps
* Discovery: Remove all targets when the scrape configuration gets empty
* PromQL: Fix a goroutine leak in the lexer/parser
* Scrape: Fix deadlock in the scrape's manager
* Scrape: Scrape targets at fixed intervals even after Prometheus restarts
* TSDB: Support restored snapshots including the head properly
* TSDB: Repair WAL when the last record in a segment is torn
- Update to 2.5.0
+ Group targets by scrape config instead of job name
+ Marathon SD: Various changes to adapt to Marathon 1.5+
+ Discovery: Split prometheus_sd_discovered_targets metric by scrape and
notify (Alertmanager SD) as well as by section in the respective configuration
+ Enhancements:
* Support s390x platform for Linux
* API: Add prometheus_api_remote_read_queries metric tracking currently
executed or waiting remote read API requests
* Remote Read: Add prometheus_remote_storage_remote_read_queries metric
tracking currently in-flight remote read queries
* Remote Read: Reduced memory usage
* Discovery: Add prometheus_sd_discovered_targets,
prometheus_sd_received_updates_total, prometheus_sd_updates_delayed_total,
and prometheus_sd_updates_total metrics for discovery subsystem
* Discovery: Improve performance of previously slow updates of changes of targets
* Kubernetes SD: Add extended metrics
* OpenStack SD: Support discovering instances from all projects
* OpenStack SD: Discover all interfaces
* OpenStack SD: Support tls_config for the used HTTP client
* Triton SD: Add ability to filter triton_sd targets by pre-defined groups
* Web UI: Avoid browser spell-checking in expression field
* Web UI: Add scrape duration and last evaluation time in targets and rules pages
* Web UI: Improve rule view by wrapping lines
* Rules: Error out at load time for invalid templates, rather than at evaluation time
+ Bug Fixes:
* Change max/min over_time to handle NaNs properly
* Check label name for count_values PromQL function
* Ensure that vectors and matrices do not contain identical label-sets
- Update to 2.4.3
+ Bug Fixes:
[BUGFIX] Fix panic when using custom EC2 API for SD #4672
[BUGFIX] Fix panic when Zookeeper SD cannot connect to servers #4669
[BUGFIX] Make the skip_head an optional parameter for snapshot API #4674
- Update to 2.4.2
+ Bug Fixes:
[BUGFIX] Handle WAL corruptions properly prometheus/tsdb#389
[BUGFIX] Handle WAL migrations correctly on Windows prometheus/tsdb#392
- Update to 2.4.1
+ New TSDB metrics
+ Bug Fixes:
Render UI correctly for Windows
- Update to 2.4.0
+ The WAL implementation has been re-written so the storage is not forward
compatible. Prometheus 2.3 storage will work on 2.4 but not vice-versa
+ Reduce remote write default retries
+ Remove /heap endpoint
+ Features:
* Persist alert 'for' state across restarts
* Add API providing per target metric metadata
* Add API providing recording and alerting rules
+ Enhancements:
* Brand new WAL implementation for TSDB. Forwards incompatible with previous WAL.
* Show rule evaluation errors in UI
* Throttle resends of alerts to Alertmanager
* Send EndsAt along with the alert to Alertmanager
* Limit the samples returned by remote read endpoint
* Limit the data read in through remote read
* Coalesce identical SD configuations
* promtool: Add new commands for debugging and querying
* Update console examples for node_exporter v0.16.0
* Optimize PromQL aggregations
* Remote read: Add Offset to hints
* consul_sd: Add support for ServiceMeta field
* ec2_sd: Maintain order of subnet_id label
* ec2_sd: Add support for custom endpoint to support EC2 compliant APIs
* ec2_sd: Add instance_owner label
* azure_sd: Add support for VMSS discovery and multiple environments
* gce_sd: Add instance_id label
* Forbid rule-abiding robots from indexing
* Log virtual memory limits on startup
+ Bug Fixes:
* Wait for service discovery to stop before exiting
* Render SD configs properly
* Only add LookbackDelta to vector selectors
* ec2_sd: Handle panic-ing nil pointer
* consul_sd: Stop leaking connections
* Use templated labels also to identify alerts
* Reduce floating point errors in stddev and related functions
* Log errors while encoding responses
- Update to 2.3.2
+ Bug Fixes:
* Fix various tsdb bugs
* Reorder startup and shutdown to prevent panics.
* Exit with non-zero code on error
* discovery/kubernetes/ingress: fix scheme discovery
* Fix race in zookeeper sd
* Better timeout handling in promql
* Propogate errors when selecting series from the tsdb
- Update to 2.3.1
+ Bug Fixes:
* Avoid infinite loop on duplicate NaN values.
* Fix nil pointer deference when using various API endpoints
* config: set target group source index during unmarshalling
* discovery/file: fix logging
* kubernetes_sd: fix namespace filtering
* web: restore old path prefix behavior
* web: remove security headers added in 2.3.0
- Update to 2.3.0
+ marathon_sd: use auth_token and auth_token_file for token-based authentication
instead of bearer_token and bearer_token_file respectively
+ Metric names for HTTP server metrics changed
+ Features:
* Add query commands to promtool
* Add security headers to HTTP server responses
* Pass query hints via remote read API
* Basic auth passwords can now be configured via file across all configuration
+ Enhancements:
* Optimise PromQL and API serialization for memory usage and allocations
* Limit number of dropped targets in web UI
* Consul and EC2 service discovery allow using server-side filtering for performance improvement
* Add advanced filtering configuration to EC2 service discovery
* marathon_sd: adds support for basic and bearer authentication, plus all
other common HTTP client options (TLS config, proxy URL, etc.)
* Provide machine type metadata and labels in GCE service discovery
* Add pod controller kind and name to Kubernetes service discovery data
* Move TSDB to flock-based log file that works with Docker containers
+ Bug Fixes:
* Properly propagate storage errors in PromQL
* Fix path prefix for web pages
* Fix goroutine leak in Consul service discovery
* Fix races in scrape manager
* Fix OOM for very large k in PromQL topk() queries
* Make remote write more resilient to unavailable receivers
* Make remote write shutdown cleanly
* Don't leak files on errors in TSDB's tombstone cleanup
* Unary minus expressions now removes the metric name from results
* Fix bug that lead to wrong amount of samples considered for time range expressions
- Update to 2.2.1
+ Bug Fixes:
* Fix data loss in TSDB on compaction
* Correctly stop timer in remote-write path
* Fix deadlock triggered by loading targets page
* Fix incorrect buffering of samples on range selection queries
* Handle large index files on windows properly
- Update to 2.2.0
+ This release introduces improvements to the storage format and fixes a
regression introduced in 2.1. As a result Prometheus servers upgraded
to 2.2 cannot be downgraded to a lower version anymore!
+ Rename file SD mtime metric
+ Send target update on empty pod IP in Kubernetes SD
+ Features:
* Add API endpoint for flags.
* Add API endpoint for dropped targets.
* Display annotations on alerts page.
* Add option to skip head data when taking snapshots
+ Enhancements:
* Federation performance improvement.
* Read bearer token file on every scrape.
* Improve typeahead on /graph page.
* Change rule file formatting.
* Set consul server default to localhost:8500.
* Add dropped Alertmanagers to API info endpoint.
* Add OS type meta label to Azure SD.
* Validate required fields in SD configuration.
+ Bug Fixes:
* Prevent stack overflow on deep recursion in TSDB.
* Correctly read offsets in index files that are greater than 4GB.
* Fix scraping behavior for empty labels.
* Drop metric name for bool modifier.
* Fix races in discovery.
* Fix Kubernetes endpoints SD for empty subsets.
* Throttle updates from SD providers, which caused increased CPU usage and allocations.
* Fix TSDB block reload issue.
* Fix PromQL printing of empty without().
* Don't reset FiredAt for inactive alerts.
* Fix erroneous file version changes and repair existing data.
- remove pr-3174.patch, has been fixed in https://github.com/prometheus/prometheus/pull/3517
- spec: fix directory for fdupes
- Update to 2.1.0:
+ Features:
* New Service Discovery UI showing labels before and after relabelling.
* New Admin APIs added to v1 to delete, snapshot and remove tombstones.
+ Enhancements:
* The graph UI autcomplete now includes your previous queries.
* Federation is now much faster for large numbers of series.
* Added new metrics to measure rule timings.
* Rule evaluation times added to the rules UI.
* Added metrics to measure modified time of file SD files.
* Kubernetes SD now includes POD UID in discovery metadata.
* The Query APIs now return optional stats on query execution times.
* The index now no longer has the 4GiB size limit and is also smaller.
+ Bug Fixes:
* Remote read read_recent option is now false by default.
* Pass the right configuration to each Alertmanager (AM) when using
multiple AM configs.
* Fix not-matchers not selecting series with labels unset.
* tsdb: Fix occasional panic in head block.
* tsdb: Close files before deletion to fix retention issues on Windows
and NFS.
* tsdb: Cleanup and do not retry failing compactions.
* tsdb: Close WAL while shutting down.
- Update to final v2.0.0 release (bsc#1067341):
+ Compared to version 1.x, this release includes a completely rewritten
storage engine, huge performance improvements, but also many backwards
incompatible changes. For more information, read the announcement blog
post and migration guide:
- https://prometheus.io/blog/2017/11/08/announcing-prometheus-2-0/
- https://prometheus.io/docs/prometheus/2.0/migration/
+ Changes:
* Completely rewritten storage layer, with WAL. This is not backwards
compatible with 1.x storage, and many flags have changed/disappeared.
* New staleness behavior. Series now marked stale after target scrapes
no longer return them, and soon after targets disappear from service
discovery.
* Rules files use YAML syntax now. Conversion tool added to promtool.
* Removed count_scalar, drop_common_labels functions and keep_common
modifier from PromQL.
* Rewritten exposition format parser with much higher performance. The
Protobuf exposition format is no longer supported.
* Example console templates updated for new storage and metrics names.
Examples other than node exporter and Prometheus removed.
* Admin and lifecycle APIs now disabled by default, can be reenabled via
flags
* Flags switched to using Kingpin, all flags are now --flagname rather
than -flagname.
+ Features:
* Remote read can be configured to not read data which is available
locally. This is enabled by default.
* Rules can be grouped now. Rules within a rule group are executed
sequentially.
* Added experimental GRPC apis
* Add timestamp() function to PromQL.
+ Enhancements:
* Remove remote read from the query path if no remote storage is
configured.
* Bump Consul HTTP client timeout to not match the Consul SD watch
timeout.
* Go-conntrack added to provide HTTP connection metrics.
+ Bug Fixes:
* Fix connection leak in Consul SD.
- Update to v2.0.0-rc.3 (bsc#1067030):
+ Enhancements:
* Remove remote read from the query path if no remote storage
is configured.
* Bump Consul HTTP client timeout to not match the Consul SD
watch timeout.
* Bump up a too small max block duration to the min block
duration instead of returning an error.
+ Bug Fixes:
* Avoid needless padding of 4 zero bytes in index files.
* Delete old blocks during reload (necessary on MS Windows).
* Fix regression of alert rules state loss on config reload.
* Serialize background WAL operations to avoid data races.
+ This also contains bugfixes and remote-storage features from
the 1.8 branch.
- Update to v2.0.0-rc.2 (bsc#1065377):
+ Enhancements:
* Handle WAL segments with corrupted header gracefully
* Stabilize memory usage during WAL replay
+ Changes:
* Prefix all storage metrics with prometheus_
+ Bug Fixes:
* Correctly handle label removal in remote read
* Fix chunk misalignment causing out-of-order samples
* Fix connection leak in Consul SD
* Handle invalid chunk dereferences gracefully
* Prevent potential deadlock during failing querier construction
- Update to v2.0.0-rc.1:
+ Features/Enhancements:
* Added a warning for time-drift between the browser and the
prometheus-server.
* Much faster WAL read-back on restart.
+ Bug Fixes:
* Fixed Remote-read to not drop the first series.
* Validate recording-rule names.
* Fix several races.
* Only close blocks if there are no iterators accessing it.
- Refresh 0002-Default-settings.patch
- Update to v2.0.0-rc.0:
+ Numerous changes to the new storage layer, the main changes being:
* Remove `count_scalar`, `keep_common` and `drop_common_labels` functions
* Breaking change in the index format for better consistency
* Fix panic due garbage collected mmap'ed strings
* Fix broken snapshots and admin APIs
* Send HTTP Accept header when scraping
* Use the WAL flush interval passed instead of the hardcoded value
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
- Refresh 0002-Default-settings.patch
- Fix loopback address for REST API gateway (bsc#1059462)
+ Added pr-3174.patch
- Update to v2.0.0-beta.5:
+ Bug Fixes:
* Remove deadlock on startup when restoring WAL
* Fix semantical races resulting in invalid persisted files
* Correctly read back WAL in certain edge cases
* Prevent crashes caused by changing metric representations in target's
/metrics
+ Enhancements:
* Overall memory usage reduction
* Serve debugging endpoints while TSDB is loading
* Healthy endpoint correctly reflects liveness during startup
* Switch to consistent usage of go-kit/log
+ This release may have issues with files written by previous beta releases
- Refresh 0002-Default-settings.patch to apply cleanly again
- Update to v2.0.0-beta.4:
+ Numerous changes to the new storage layer, the main changes being:
* Single, compacted write ahead log
* Single in-memory block with garbage collection
* Improve compression of index files
* Cache series dropped via metric_relabel_configs
* Pool byte buffers for scraping
+ Overall the changes achieve a baseline reduction in memory consumption
and reduce peak memory usage by 30-40% compared to 2.0.0-beta.2
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
- Update to v2.0.0-beta.2
+ vendor: update prometheus/tsdb (Fix panic caused by 0 division)
- Update to v2.0.0-beta.1
+ Includes a new storage layer, which reduces bottlenecks and shows
considerable performance improvements, but does not work with old v1
storage data.
- Refresh 0002-Default-settings.patch to apply cleanly again
- Add systemd unit file and default config (/etc/prometheus/prometheus.yml)
- Run as unprivileged "prometheus" user
- Default to /etc/prometheus for config and /var/lib/prometheus for metrics
+ Added 0002-Default-settings.patch
- Add BuildRequires: glibc-devel-static to fix aarch64 build
- Initial version
OBS-URL: https://build.opensuse.org/request/show/720641
OBS-URL: https://build.opensuse.org/package/show/server:monitoring/golang-github-prometheus-prometheus?expand=0&rev=17
2019-08-07 10:15:50 +02:00
-------------------------------------------------------------------
Fri Aug 2 11:22:25 UTC 2019 - Jan Fajerski <jan.fajerski@suse.com>
- Add network-online (Wants and After) dependency to systemd unit bsc#1143913
2019-08-02 13:20:08 +02:00
-------------------------------------------------------------------
Wed Jul 31 10:46:13 UTC 2019 - Andreas Schneider <asn@cryptomilk.org>
- Build with PIE
Accepting request 720641 from home:jfajerski:branches:server:monitoring
- Add network-online (Wants and After) dependency to systemd unit bsc#1143913
- Build with PIE
+ Added 0004-prometheus-buildmode-pie.patch
- Only package required files (reduces rpm size by 4 MB)
- Add sysconfig file
- Add firewall config file
- Use variables for defining user and group
- Add support for Uyuni/SUSE Manager service discovery
+ Added 0003-Add-Uyuni-service-discovery.patch
- readded _service file removed in error.
- Update to 2.11.1
+ Bug Fix:
* Fix potential panic when prometheus is watching multiple zookeeper paths.
- Update to 2.11.0
+ Bug Fix:
* resolve race condition in maxGauge.
* Fix ZooKeeper connection leak.
* Improved atomicity of .tmp block replacement during compaction for usual case.
* Fix "unknown series references" after clean shutdown.
* Re-calculate block size when calling block.Delete.
* Fix unsafe snapshots with head block.
* prometheus_tsdb_compactions_failed_total is now incremented on any compaction failure.
+ Changes:
* Remove max_retries from queue_config (it has been unused since rewriting remote-write to utilize the write-ahead-log)
* The meta file BlockStats no longer holds size information. This is now dynamically calculated and kept in memory. It also includes the meta file size which was not included before
* Renamed metric from prometheus_tsdb_wal_reader_corruption_errors to prometheus_tsdb_wal_reader_corruption_errors_total
+ Features:
* Add option to use Alertmanager API v2.
* Added humanizePercentage function for templates.
* Include InitContainers in Kubernetes Service Discovery.
* Provide option to compress WAL records using Snappy.
+ Enhancements:
* Create new clean segment when starting the WAL.
* Reduce allocations in PromQL aggregations.
* Add storage warnings to LabelValues and LabelNames API results.
* Add prometheus_http_requests_total metric.
* Enable openbsd/arm build.
* Remote-write allocation improvements.
* Query performance improvement: Efficient iteration and search in HashForLabels and HashWithoutLabels.
* Allow injection of arbitrary headers in promtool.
* Allow passing external_labels in alert unit tests groups.
* Allows globs for rules when unit testing.
* Improved postings intersection matching.
* Reduced disk usage for WAL for small setups.
* Optimize queries using regexp for set lookups.
- rebase patch002-Default-settings.patch
- Update to 2.10.0:
+ Bug Fixes:
* TSDB: Don't panic when running out of disk space and recover nicely from the condition
* TSDB: Correctly handle empty labels.
* TSDB: Don't crash on an unknown tombstone reference.
* Storage/remote: Remove queue-manager specific metrics if queue no longer exists.
* PromQL: Correctly display {__name__="a"}.
* Discovery/kubernetes: Use service rather than ingress as the name for the service workqueue.
* Discovery/azure: Don't panic on a VM with a public IP.
* Web: Fixed Content-Type for js and css instead of using /etc/mime.types.
* API: Encode alert values as string to correctly represent Inf/NaN.
+ Features:
* Template expansion: Make external labels available as $externalLabels in alert and console template expansion.
* TSDB: Add prometheus_tsdb_wal_segment_current metric for the WAL segment index that TSDB is currently writing to. tsdb
* Scrape: Add scrape_series_added per-scrape metric. #5546
+ Enhancements
* Discovery/kubernetes: Add labels __meta_kubernetes_endpoint_node_name and __meta_kubernetes_endpoint_hostname.
* Discovery/azure: Add label __meta_azure_machine_public_ip.
* TSDB: Simplify mergedPostings.Seek, resulting in better performance if there are many posting lists. tsdb
* Log filesystem type on startup.
* Cmd/promtool: Use POST requests for Query and QueryRange. client_golang
* Web: Sort alerts by group name.
* Console templates: Add convenience variables $rawParams, $params, $path.
- Upadte to 2.9.2
+ Bug Fixes:
* Make sure subquery range is taken into account for selection
* Exhaust every request body before closing it
* Cmd/promtool: return errors from rule evaluations
* Remote Storage: string interner should not panic in release
* Fix memory allocation regression in mergedPostings.Seek tsdb
- Update to 2.9.1
+ Bug Fixes:
* Discovery/kubernetes: fix missing label sanitization
* Remote_write: Prevent reshard concurrent with calling stop
- Update to 2.9.0
+ Feature:
* Add honor_timestamps scrape option.
+ Enhancements:
* Update Consul to support catalog.ServiceMultipleTags.
* Discovery/kubernetes: add present labels for labels/annotations.
* OpenStack SD: Add ProjectID and UserID meta labels.
* Add GODEBUG and retention to the runtime page.
* Add support for POSTing to /series endpoint.
* Support PUT methods for Lifecycle and Admin APIs.
* Scrape: Add global jitter for HA server.
* Check for cancellation on every step of a range evaluation.
* String interning for labels & values in the remote_write path.
* Don't lose the scrape cache on a failed scrape.
* Reload cert files from disk automatically. common
* Use fixed length millisecond timestamp format for logs. common
* Performance improvements for postings.
Bug Fixes:
* Remote Write: fix checkpoint reading.
* Check if label value is valid when unmarshaling external labels from YAML.
* Promparse: sort all labels when parsing.
* Reload rules: copy state on both name and labels.
* Exponentation operator to drop metric name in result of operation.
* Config: resolve more file paths.
* Promtool: resolve relative paths in alert test files.
* Set TLSHandshakeTimeout in HTTP transport. common
* Use fsync to be more resilient to machine crashes.
* Keep series that are still in WAL in checkpoints.
- Update to 2.8.1
+ Bug Fixes
* Display the job labels in /targets which was removed accidentally
- Update to 2.8.0
+ Change:
* This release uses Write-Ahead Logging (WAL) for the remote_write API. This currently causes a slight increase in memory usage, which will be addressed in future releases.
* Default time retention is used only when no size based retention is specified. These are flags where time retention is specified by the flag --storage.tsdb.retention and size retention by --storage.tsdb.retention.size.
* prometheus_tsdb_storage_blocks_bytes_total is now prometheus_tsdb_storage_blocks_bytes.
+ Feature:
* (EXPERIMENTAL) Time overlapping blocks are now allowed; vertical compaction and vertical query merge. It is an optional feature which is controlled by the --storage.tsdb.allow-overlapping-blocks flag, disabled by default.
+ Enhancements:
* Use the WAL for remote_write API.
* Query performance improvements.
* UI enhancements with upgrade to Bootstrap 4.
* Reduce time that Alertmanagers are in flux when reloaded.
* Limit number of metrics displayed on UI to 10000.
* (1) Remember All/Unhealthy choice on target-overview when reloading page. (2) Resize text-input area on Graph page on mouseclick.
* In histogram_quantile merge buckets with equivalent le values.
* Show list of offending labels in the error message in many-to-many scenarios.
* Show Storage Retention criteria in effect on /status page.
+ Bug Fixes:
+ Fix sorting of rule groups.
+ Fix support for password_file and bearer_token_file in Kubernetes SD.
+ Scrape: catch errors when creating HTTP clients
+ Adds new metrics:
prometheus_target_scrape_pools_total
prometheus_target_scrape_pools_failed_total
prometheus_target_scrape_pool_reloads_total
prometheus_target_scrape_pool_reloads_failed_total
+ Fix panic when aggregator param is not a literal.
- fix spec file: actually ship promtool
- Update to 2.7.1:
+ Bug Fixes:
* Fix a Stored DOM XSS vulnerability with query history (boo#1124610)
* prometheus_rule_group_last_duration_seconds now reports seconds instead of nanoseconds
* Make sure the targets are consistently sorted in the targets page
- Update to 2.7.0:
+ cli flag depreacted: storage.tsdb.retention use storage.tsdb.retention.time
instead; depreacted flag will be removed in 3.0
+ Features:
* Add subqueries to PromQL
* Add support for disk size based retention. Note that we don't
consider the WAL size which could be significant and the time
based retention policy also applies (experimental)
* Add CORS origin flag
+ Bug Fixes:
* Don't depend on given order when comparing samples in alert unit testing
* Make sure the retention period doesn't overflow
* Don't generate blocks with no samples
- Update to 2.6.0:
+ Remove default flags from the container's entrypoint, run Prometheus from
/etc/prometheus and symlink the storage directory to /etc/prometheus/data
+ Promtool: Remove the update command
+ Features:
* Add JSON log format via the --log.format flag
* API: Add /api/v1/labels endpoint to get all label names
* Web: Allow setting the page's title via the --web.ui-title flag
+ Enhancements:
* Add prometheus_tsdb_lowest_timestamp_seconds, prometheus_tsdb_head_min_time_seconds
and prometheus_tsdb_head_max_time_seconds metrics
* Add rule_group_last_evaluation_timestamp_seconds metric
* Add prometheus_template_text_expansion_failures_total and prometheus_template_text_expansions_total metrics
* Set consistent User-Agent header in outgoing requests
* Azure SD: Error out at load time when authentication parameters are missing
* EC2 SD: Add the machine's private DNS name to the discovery metadata
* EC2 SD: Add the operating system's platform to the discovery metadata
* Kubernetes SD: Add the pod's phase to the discovery metadata
* Kubernetes SD: Log Kubernetes messages
* Promtool: Collect CPU and trace profiles
* Promtool: Support writing output as JSON
* Remote Read: Return available data if remote read fails partially
* Remote Write: Improve queue performance
* Remote Write: Add min_shards parameter to set the minimum number of shards
* TSDB: Improve WAL reading
* TSDB: Memory improvements
* Web: Log stack traces on panic
* Web UI: Add copy to clipboard button for configuration
* Web UI: Support console queries at specific times
* Web UI: group targets by job then instance
+ Bug Fixes:
* Deduplicate handler labels for HTTP metrics
* Fix leaked queriers causing shutdowns to hang
* Fix configuration loading panics on nil pointer slice elements
* API: Correctly skip mismatching targets on /api/v1/targets/metadata
* API: Better rounding for incoming query timestamps
* Discovery: Remove all targets when the scrape configuration gets empty
* PromQL: Fix a goroutine leak in the lexer/parser
* Scrape: Fix deadlock in the scrape's manager
* Scrape: Scrape targets at fixed intervals even after Prometheus restarts
* TSDB: Support restored snapshots including the head properly
* TSDB: Repair WAL when the last record in a segment is torn
- Update to 2.5.0
+ Group targets by scrape config instead of job name
+ Marathon SD: Various changes to adapt to Marathon 1.5+
+ Discovery: Split prometheus_sd_discovered_targets metric by scrape and
notify (Alertmanager SD) as well as by section in the respective configuration
+ Enhancements:
* Support s390x platform for Linux
* API: Add prometheus_api_remote_read_queries metric tracking currently
executed or waiting remote read API requests
* Remote Read: Add prometheus_remote_storage_remote_read_queries metric
tracking currently in-flight remote read queries
* Remote Read: Reduced memory usage
* Discovery: Add prometheus_sd_discovered_targets,
prometheus_sd_received_updates_total, prometheus_sd_updates_delayed_total,
and prometheus_sd_updates_total metrics for discovery subsystem
* Discovery: Improve performance of previously slow updates of changes of targets
* Kubernetes SD: Add extended metrics
* OpenStack SD: Support discovering instances from all projects
* OpenStack SD: Discover all interfaces
* OpenStack SD: Support tls_config for the used HTTP client
* Triton SD: Add ability to filter triton_sd targets by pre-defined groups
* Web UI: Avoid browser spell-checking in expression field
* Web UI: Add scrape duration and last evaluation time in targets and rules pages
* Web UI: Improve rule view by wrapping lines
* Rules: Error out at load time for invalid templates, rather than at evaluation time
+ Bug Fixes:
* Change max/min over_time to handle NaNs properly
* Check label name for count_values PromQL function
* Ensure that vectors and matrices do not contain identical label-sets
- Update to 2.4.3
+ Bug Fixes:
[BUGFIX] Fix panic when using custom EC2 API for SD #4672
[BUGFIX] Fix panic when Zookeeper SD cannot connect to servers #4669
[BUGFIX] Make the skip_head an optional parameter for snapshot API #4674
- Update to 2.4.2
+ Bug Fixes:
[BUGFIX] Handle WAL corruptions properly prometheus/tsdb#389
[BUGFIX] Handle WAL migrations correctly on Windows prometheus/tsdb#392
- Update to 2.4.1
+ New TSDB metrics
+ Bug Fixes:
Render UI correctly for Windows
- Update to 2.4.0
+ The WAL implementation has been re-written so the storage is not forward
compatible. Prometheus 2.3 storage will work on 2.4 but not vice-versa
+ Reduce remote write default retries
+ Remove /heap endpoint
+ Features:
* Persist alert 'for' state across restarts
* Add API providing per target metric metadata
* Add API providing recording and alerting rules
+ Enhancements:
* Brand new WAL implementation for TSDB. Forwards incompatible with previous WAL.
* Show rule evaluation errors in UI
* Throttle resends of alerts to Alertmanager
* Send EndsAt along with the alert to Alertmanager
* Limit the samples returned by remote read endpoint
* Limit the data read in through remote read
* Coalesce identical SD configuations
* promtool: Add new commands for debugging and querying
* Update console examples for node_exporter v0.16.0
* Optimize PromQL aggregations
* Remote read: Add Offset to hints
* consul_sd: Add support for ServiceMeta field
* ec2_sd: Maintain order of subnet_id label
* ec2_sd: Add support for custom endpoint to support EC2 compliant APIs
* ec2_sd: Add instance_owner label
* azure_sd: Add support for VMSS discovery and multiple environments
* gce_sd: Add instance_id label
* Forbid rule-abiding robots from indexing
* Log virtual memory limits on startup
+ Bug Fixes:
* Wait for service discovery to stop before exiting
* Render SD configs properly
* Only add LookbackDelta to vector selectors
* ec2_sd: Handle panic-ing nil pointer
* consul_sd: Stop leaking connections
* Use templated labels also to identify alerts
* Reduce floating point errors in stddev and related functions
* Log errors while encoding responses
- Update to 2.3.2
+ Bug Fixes:
* Fix various tsdb bugs
* Reorder startup and shutdown to prevent panics.
* Exit with non-zero code on error
* discovery/kubernetes/ingress: fix scheme discovery
* Fix race in zookeeper sd
* Better timeout handling in promql
* Propogate errors when selecting series from the tsdb
- Update to 2.3.1
+ Bug Fixes:
* Avoid infinite loop on duplicate NaN values.
* Fix nil pointer deference when using various API endpoints
* config: set target group source index during unmarshalling
* discovery/file: fix logging
* kubernetes_sd: fix namespace filtering
* web: restore old path prefix behavior
* web: remove security headers added in 2.3.0
- Update to 2.3.0
+ marathon_sd: use auth_token and auth_token_file for token-based authentication
instead of bearer_token and bearer_token_file respectively
+ Metric names for HTTP server metrics changed
+ Features:
* Add query commands to promtool
* Add security headers to HTTP server responses
* Pass query hints via remote read API
* Basic auth passwords can now be configured via file across all configuration
+ Enhancements:
* Optimise PromQL and API serialization for memory usage and allocations
* Limit number of dropped targets in web UI
* Consul and EC2 service discovery allow using server-side filtering for performance improvement
* Add advanced filtering configuration to EC2 service discovery
* marathon_sd: adds support for basic and bearer authentication, plus all
other common HTTP client options (TLS config, proxy URL, etc.)
* Provide machine type metadata and labels in GCE service discovery
* Add pod controller kind and name to Kubernetes service discovery data
* Move TSDB to flock-based log file that works with Docker containers
+ Bug Fixes:
* Properly propagate storage errors in PromQL
* Fix path prefix for web pages
* Fix goroutine leak in Consul service discovery
* Fix races in scrape manager
* Fix OOM for very large k in PromQL topk() queries
* Make remote write more resilient to unavailable receivers
* Make remote write shutdown cleanly
* Don't leak files on errors in TSDB's tombstone cleanup
* Unary minus expressions now removes the metric name from results
* Fix bug that lead to wrong amount of samples considered for time range expressions
- Update to 2.2.1
+ Bug Fixes:
* Fix data loss in TSDB on compaction
* Correctly stop timer in remote-write path
* Fix deadlock triggered by loading targets page
* Fix incorrect buffering of samples on range selection queries
* Handle large index files on windows properly
- Update to 2.2.0
+ This release introduces improvements to the storage format and fixes a
regression introduced in 2.1. As a result Prometheus servers upgraded
to 2.2 cannot be downgraded to a lower version anymore!
+ Rename file SD mtime metric
+ Send target update on empty pod IP in Kubernetes SD
+ Features:
* Add API endpoint for flags.
* Add API endpoint for dropped targets.
* Display annotations on alerts page.
* Add option to skip head data when taking snapshots
+ Enhancements:
* Federation performance improvement.
* Read bearer token file on every scrape.
* Improve typeahead on /graph page.
* Change rule file formatting.
* Set consul server default to localhost:8500.
* Add dropped Alertmanagers to API info endpoint.
* Add OS type meta label to Azure SD.
* Validate required fields in SD configuration.
+ Bug Fixes:
* Prevent stack overflow on deep recursion in TSDB.
* Correctly read offsets in index files that are greater than 4GB.
* Fix scraping behavior for empty labels.
* Drop metric name for bool modifier.
* Fix races in discovery.
* Fix Kubernetes endpoints SD for empty subsets.
* Throttle updates from SD providers, which caused increased CPU usage and allocations.
* Fix TSDB block reload issue.
* Fix PromQL printing of empty without().
* Don't reset FiredAt for inactive alerts.
* Fix erroneous file version changes and repair existing data.
- remove pr-3174.patch, has been fixed in https://github.com/prometheus/prometheus/pull/3517
- spec: fix directory for fdupes
- Update to 2.1.0:
+ Features:
* New Service Discovery UI showing labels before and after relabelling.
* New Admin APIs added to v1 to delete, snapshot and remove tombstones.
+ Enhancements:
* The graph UI autcomplete now includes your previous queries.
* Federation is now much faster for large numbers of series.
* Added new metrics to measure rule timings.
* Rule evaluation times added to the rules UI.
* Added metrics to measure modified time of file SD files.
* Kubernetes SD now includes POD UID in discovery metadata.
* The Query APIs now return optional stats on query execution times.
* The index now no longer has the 4GiB size limit and is also smaller.
+ Bug Fixes:
* Remote read read_recent option is now false by default.
* Pass the right configuration to each Alertmanager (AM) when using
multiple AM configs.
* Fix not-matchers not selecting series with labels unset.
* tsdb: Fix occasional panic in head block.
* tsdb: Close files before deletion to fix retention issues on Windows
and NFS.
* tsdb: Cleanup and do not retry failing compactions.
* tsdb: Close WAL while shutting down.
- Update to final v2.0.0 release (bsc#1067341):
+ Compared to version 1.x, this release includes a completely rewritten
storage engine, huge performance improvements, but also many backwards
incompatible changes. For more information, read the announcement blog
post and migration guide:
- https://prometheus.io/blog/2017/11/08/announcing-prometheus-2-0/
- https://prometheus.io/docs/prometheus/2.0/migration/
+ Changes:
* Completely rewritten storage layer, with WAL. This is not backwards
compatible with 1.x storage, and many flags have changed/disappeared.
* New staleness behavior. Series now marked stale after target scrapes
no longer return them, and soon after targets disappear from service
discovery.
* Rules files use YAML syntax now. Conversion tool added to promtool.
* Removed count_scalar, drop_common_labels functions and keep_common
modifier from PromQL.
* Rewritten exposition format parser with much higher performance. The
Protobuf exposition format is no longer supported.
* Example console templates updated for new storage and metrics names.
Examples other than node exporter and Prometheus removed.
* Admin and lifecycle APIs now disabled by default, can be reenabled via
flags
* Flags switched to using Kingpin, all flags are now --flagname rather
than -flagname.
+ Features:
* Remote read can be configured to not read data which is available
locally. This is enabled by default.
* Rules can be grouped now. Rules within a rule group are executed
sequentially.
* Added experimental GRPC apis
* Add timestamp() function to PromQL.
+ Enhancements:
* Remove remote read from the query path if no remote storage is
configured.
* Bump Consul HTTP client timeout to not match the Consul SD watch
timeout.
* Go-conntrack added to provide HTTP connection metrics.
+ Bug Fixes:
* Fix connection leak in Consul SD.
- Update to v2.0.0-rc.3 (bsc#1067030):
+ Enhancements:
* Remove remote read from the query path if no remote storage
is configured.
* Bump Consul HTTP client timeout to not match the Consul SD
watch timeout.
* Bump up a too small max block duration to the min block
duration instead of returning an error.
+ Bug Fixes:
* Avoid needless padding of 4 zero bytes in index files.
* Delete old blocks during reload (necessary on MS Windows).
* Fix regression of alert rules state loss on config reload.
* Serialize background WAL operations to avoid data races.
+ This also contains bugfixes and remote-storage features from
the 1.8 branch.
- Update to v2.0.0-rc.2 (bsc#1065377):
+ Enhancements:
* Handle WAL segments with corrupted header gracefully
* Stabilize memory usage during WAL replay
+ Changes:
* Prefix all storage metrics with prometheus_
+ Bug Fixes:
* Correctly handle label removal in remote read
* Fix chunk misalignment causing out-of-order samples
* Fix connection leak in Consul SD
* Handle invalid chunk dereferences gracefully
* Prevent potential deadlock during failing querier construction
- Update to v2.0.0-rc.1:
+ Features/Enhancements:
* Added a warning for time-drift between the browser and the
prometheus-server.
* Much faster WAL read-back on restart.
+ Bug Fixes:
* Fixed Remote-read to not drop the first series.
* Validate recording-rule names.
* Fix several races.
* Only close blocks if there are no iterators accessing it.
- Refresh 0002-Default-settings.patch
- Update to v2.0.0-rc.0:
+ Numerous changes to the new storage layer, the main changes being:
* Remove `count_scalar`, `keep_common` and `drop_common_labels` functions
* Breaking change in the index format for better consistency
* Fix panic due garbage collected mmap'ed strings
* Fix broken snapshots and admin APIs
* Send HTTP Accept header when scraping
* Use the WAL flush interval passed instead of the hardcoded value
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
- Refresh 0002-Default-settings.patch
- Fix loopback address for REST API gateway (bsc#1059462)
+ Added pr-3174.patch
- Update to v2.0.0-beta.5:
+ Bug Fixes:
* Remove deadlock on startup when restoring WAL
* Fix semantical races resulting in invalid persisted files
* Correctly read back WAL in certain edge cases
* Prevent crashes caused by changing metric representations in target's
/metrics
+ Enhancements:
* Overall memory usage reduction
* Serve debugging endpoints while TSDB is loading
* Healthy endpoint correctly reflects liveness during startup
* Switch to consistent usage of go-kit/log
+ This release may have issues with files written by previous beta releases
- Refresh 0002-Default-settings.patch to apply cleanly again
- Update to v2.0.0-beta.4:
+ Numerous changes to the new storage layer, the main changes being:
* Single, compacted write ahead log
* Single in-memory block with garbage collection
* Improve compression of index files
* Cache series dropped via metric_relabel_configs
* Pool byte buffers for scraping
+ Overall the changes achieve a baseline reduction in memory consumption
and reduce peak memory usage by 30-40% compared to 2.0.0-beta.2
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
- Update to v2.0.0-beta.2
+ vendor: update prometheus/tsdb (Fix panic caused by 0 division)
- Update to v2.0.0-beta.1
+ Includes a new storage layer, which reduces bottlenecks and shows
considerable performance improvements, but does not work with old v1
storage data.
- Refresh 0002-Default-settings.patch to apply cleanly again
- Add systemd unit file and default config (/etc/prometheus/prometheus.yml)
- Run as unprivileged "prometheus" user
- Default to /etc/prometheus for config and /var/lib/prometheus for metrics
+ Added 0002-Default-settings.patch
- Add BuildRequires: glibc-devel-static to fix aarch64 build
- Initial version
OBS-URL: https://build.opensuse.org/request/show/720641
OBS-URL: https://build.opensuse.org/package/show/server:monitoring/golang-github-prometheus-prometheus?expand=0&rev=17
2019-08-07 10:15:50 +02:00
+ Added 0004-prometheus-buildmode-pie.patch
2019-08-02 13:20:08 +02:00
-------------------------------------------------------------------
Wed Jul 31 06:47:55 UTC 2019 - Andreas Schneider <asn@cryptomilk.org>
- Only package required files (reduces rpm size by 4 MB)
- Add sysconfig file
- Add firewall config file
- Use variables for defining user and group
2019-07-26 13:24:49 +02:00
-------------------------------------------------------------------
Thu Jul 25 16:34:29 UTC 2019 - Joao Cavalheiro <jcavalheiro@suse.com>
- Add support for Uyuni/SUSE Manager service discovery
Accepting request 720641 from home:jfajerski:branches:server:monitoring
- Add network-online (Wants and After) dependency to systemd unit bsc#1143913
- Build with PIE
+ Added 0004-prometheus-buildmode-pie.patch
- Only package required files (reduces rpm size by 4 MB)
- Add sysconfig file
- Add firewall config file
- Use variables for defining user and group
- Add support for Uyuni/SUSE Manager service discovery
+ Added 0003-Add-Uyuni-service-discovery.patch
- readded _service file removed in error.
- Update to 2.11.1
+ Bug Fix:
* Fix potential panic when prometheus is watching multiple zookeeper paths.
- Update to 2.11.0
+ Bug Fix:
* resolve race condition in maxGauge.
* Fix ZooKeeper connection leak.
* Improved atomicity of .tmp block replacement during compaction for usual case.
* Fix "unknown series references" after clean shutdown.
* Re-calculate block size when calling block.Delete.
* Fix unsafe snapshots with head block.
* prometheus_tsdb_compactions_failed_total is now incremented on any compaction failure.
+ Changes:
* Remove max_retries from queue_config (it has been unused since rewriting remote-write to utilize the write-ahead-log)
* The meta file BlockStats no longer holds size information. This is now dynamically calculated and kept in memory. It also includes the meta file size which was not included before
* Renamed metric from prometheus_tsdb_wal_reader_corruption_errors to prometheus_tsdb_wal_reader_corruption_errors_total
+ Features:
* Add option to use Alertmanager API v2.
* Added humanizePercentage function for templates.
* Include InitContainers in Kubernetes Service Discovery.
* Provide option to compress WAL records using Snappy.
+ Enhancements:
* Create new clean segment when starting the WAL.
* Reduce allocations in PromQL aggregations.
* Add storage warnings to LabelValues and LabelNames API results.
* Add prometheus_http_requests_total metric.
* Enable openbsd/arm build.
* Remote-write allocation improvements.
* Query performance improvement: Efficient iteration and search in HashForLabels and HashWithoutLabels.
* Allow injection of arbitrary headers in promtool.
* Allow passing external_labels in alert unit tests groups.
* Allows globs for rules when unit testing.
* Improved postings intersection matching.
* Reduced disk usage for WAL for small setups.
* Optimize queries using regexp for set lookups.
- rebase patch002-Default-settings.patch
- Update to 2.10.0:
+ Bug Fixes:
* TSDB: Don't panic when running out of disk space and recover nicely from the condition
* TSDB: Correctly handle empty labels.
* TSDB: Don't crash on an unknown tombstone reference.
* Storage/remote: Remove queue-manager specific metrics if queue no longer exists.
* PromQL: Correctly display {__name__="a"}.
* Discovery/kubernetes: Use service rather than ingress as the name for the service workqueue.
* Discovery/azure: Don't panic on a VM with a public IP.
* Web: Fixed Content-Type for js and css instead of using /etc/mime.types.
* API: Encode alert values as string to correctly represent Inf/NaN.
+ Features:
* Template expansion: Make external labels available as $externalLabels in alert and console template expansion.
* TSDB: Add prometheus_tsdb_wal_segment_current metric for the WAL segment index that TSDB is currently writing to. tsdb
* Scrape: Add scrape_series_added per-scrape metric. #5546
+ Enhancements
* Discovery/kubernetes: Add labels __meta_kubernetes_endpoint_node_name and __meta_kubernetes_endpoint_hostname.
* Discovery/azure: Add label __meta_azure_machine_public_ip.
* TSDB: Simplify mergedPostings.Seek, resulting in better performance if there are many posting lists. tsdb
* Log filesystem type on startup.
* Cmd/promtool: Use POST requests for Query and QueryRange. client_golang
* Web: Sort alerts by group name.
* Console templates: Add convenience variables $rawParams, $params, $path.
- Upadte to 2.9.2
+ Bug Fixes:
* Make sure subquery range is taken into account for selection
* Exhaust every request body before closing it
* Cmd/promtool: return errors from rule evaluations
* Remote Storage: string interner should not panic in release
* Fix memory allocation regression in mergedPostings.Seek tsdb
- Update to 2.9.1
+ Bug Fixes:
* Discovery/kubernetes: fix missing label sanitization
* Remote_write: Prevent reshard concurrent with calling stop
- Update to 2.9.0
+ Feature:
* Add honor_timestamps scrape option.
+ Enhancements:
* Update Consul to support catalog.ServiceMultipleTags.
* Discovery/kubernetes: add present labels for labels/annotations.
* OpenStack SD: Add ProjectID and UserID meta labels.
* Add GODEBUG and retention to the runtime page.
* Add support for POSTing to /series endpoint.
* Support PUT methods for Lifecycle and Admin APIs.
* Scrape: Add global jitter for HA server.
* Check for cancellation on every step of a range evaluation.
* String interning for labels & values in the remote_write path.
* Don't lose the scrape cache on a failed scrape.
* Reload cert files from disk automatically. common
* Use fixed length millisecond timestamp format for logs. common
* Performance improvements for postings.
Bug Fixes:
* Remote Write: fix checkpoint reading.
* Check if label value is valid when unmarshaling external labels from YAML.
* Promparse: sort all labels when parsing.
* Reload rules: copy state on both name and labels.
* Exponentation operator to drop metric name in result of operation.
* Config: resolve more file paths.
* Promtool: resolve relative paths in alert test files.
* Set TLSHandshakeTimeout in HTTP transport. common
* Use fsync to be more resilient to machine crashes.
* Keep series that are still in WAL in checkpoints.
- Update to 2.8.1
+ Bug Fixes
* Display the job labels in /targets which was removed accidentally
- Update to 2.8.0
+ Change:
* This release uses Write-Ahead Logging (WAL) for the remote_write API. This currently causes a slight increase in memory usage, which will be addressed in future releases.
* Default time retention is used only when no size based retention is specified. These are flags where time retention is specified by the flag --storage.tsdb.retention and size retention by --storage.tsdb.retention.size.
* prometheus_tsdb_storage_blocks_bytes_total is now prometheus_tsdb_storage_blocks_bytes.
+ Feature:
* (EXPERIMENTAL) Time overlapping blocks are now allowed; vertical compaction and vertical query merge. It is an optional feature which is controlled by the --storage.tsdb.allow-overlapping-blocks flag, disabled by default.
+ Enhancements:
* Use the WAL for remote_write API.
* Query performance improvements.
* UI enhancements with upgrade to Bootstrap 4.
* Reduce time that Alertmanagers are in flux when reloaded.
* Limit number of metrics displayed on UI to 10000.
* (1) Remember All/Unhealthy choice on target-overview when reloading page. (2) Resize text-input area on Graph page on mouseclick.
* In histogram_quantile merge buckets with equivalent le values.
* Show list of offending labels in the error message in many-to-many scenarios.
* Show Storage Retention criteria in effect on /status page.
+ Bug Fixes:
+ Fix sorting of rule groups.
+ Fix support for password_file and bearer_token_file in Kubernetes SD.
+ Scrape: catch errors when creating HTTP clients
+ Adds new metrics:
prometheus_target_scrape_pools_total
prometheus_target_scrape_pools_failed_total
prometheus_target_scrape_pool_reloads_total
prometheus_target_scrape_pool_reloads_failed_total
+ Fix panic when aggregator param is not a literal.
- fix spec file: actually ship promtool
- Update to 2.7.1:
+ Bug Fixes:
* Fix a Stored DOM XSS vulnerability with query history (boo#1124610)
* prometheus_rule_group_last_duration_seconds now reports seconds instead of nanoseconds
* Make sure the targets are consistently sorted in the targets page
- Update to 2.7.0:
+ cli flag depreacted: storage.tsdb.retention use storage.tsdb.retention.time
instead; depreacted flag will be removed in 3.0
+ Features:
* Add subqueries to PromQL
* Add support for disk size based retention. Note that we don't
consider the WAL size which could be significant and the time
based retention policy also applies (experimental)
* Add CORS origin flag
+ Bug Fixes:
* Don't depend on given order when comparing samples in alert unit testing
* Make sure the retention period doesn't overflow
* Don't generate blocks with no samples
- Update to 2.6.0:
+ Remove default flags from the container's entrypoint, run Prometheus from
/etc/prometheus and symlink the storage directory to /etc/prometheus/data
+ Promtool: Remove the update command
+ Features:
* Add JSON log format via the --log.format flag
* API: Add /api/v1/labels endpoint to get all label names
* Web: Allow setting the page's title via the --web.ui-title flag
+ Enhancements:
* Add prometheus_tsdb_lowest_timestamp_seconds, prometheus_tsdb_head_min_time_seconds
and prometheus_tsdb_head_max_time_seconds metrics
* Add rule_group_last_evaluation_timestamp_seconds metric
* Add prometheus_template_text_expansion_failures_total and prometheus_template_text_expansions_total metrics
* Set consistent User-Agent header in outgoing requests
* Azure SD: Error out at load time when authentication parameters are missing
* EC2 SD: Add the machine's private DNS name to the discovery metadata
* EC2 SD: Add the operating system's platform to the discovery metadata
* Kubernetes SD: Add the pod's phase to the discovery metadata
* Kubernetes SD: Log Kubernetes messages
* Promtool: Collect CPU and trace profiles
* Promtool: Support writing output as JSON
* Remote Read: Return available data if remote read fails partially
* Remote Write: Improve queue performance
* Remote Write: Add min_shards parameter to set the minimum number of shards
* TSDB: Improve WAL reading
* TSDB: Memory improvements
* Web: Log stack traces on panic
* Web UI: Add copy to clipboard button for configuration
* Web UI: Support console queries at specific times
* Web UI: group targets by job then instance
+ Bug Fixes:
* Deduplicate handler labels for HTTP metrics
* Fix leaked queriers causing shutdowns to hang
* Fix configuration loading panics on nil pointer slice elements
* API: Correctly skip mismatching targets on /api/v1/targets/metadata
* API: Better rounding for incoming query timestamps
* Discovery: Remove all targets when the scrape configuration gets empty
* PromQL: Fix a goroutine leak in the lexer/parser
* Scrape: Fix deadlock in the scrape's manager
* Scrape: Scrape targets at fixed intervals even after Prometheus restarts
* TSDB: Support restored snapshots including the head properly
* TSDB: Repair WAL when the last record in a segment is torn
- Update to 2.5.0
+ Group targets by scrape config instead of job name
+ Marathon SD: Various changes to adapt to Marathon 1.5+
+ Discovery: Split prometheus_sd_discovered_targets metric by scrape and
notify (Alertmanager SD) as well as by section in the respective configuration
+ Enhancements:
* Support s390x platform for Linux
* API: Add prometheus_api_remote_read_queries metric tracking currently
executed or waiting remote read API requests
* Remote Read: Add prometheus_remote_storage_remote_read_queries metric
tracking currently in-flight remote read queries
* Remote Read: Reduced memory usage
* Discovery: Add prometheus_sd_discovered_targets,
prometheus_sd_received_updates_total, prometheus_sd_updates_delayed_total,
and prometheus_sd_updates_total metrics for discovery subsystem
* Discovery: Improve performance of previously slow updates of changes of targets
* Kubernetes SD: Add extended metrics
* OpenStack SD: Support discovering instances from all projects
* OpenStack SD: Discover all interfaces
* OpenStack SD: Support tls_config for the used HTTP client
* Triton SD: Add ability to filter triton_sd targets by pre-defined groups
* Web UI: Avoid browser spell-checking in expression field
* Web UI: Add scrape duration and last evaluation time in targets and rules pages
* Web UI: Improve rule view by wrapping lines
* Rules: Error out at load time for invalid templates, rather than at evaluation time
+ Bug Fixes:
* Change max/min over_time to handle NaNs properly
* Check label name for count_values PromQL function
* Ensure that vectors and matrices do not contain identical label-sets
- Update to 2.4.3
+ Bug Fixes:
[BUGFIX] Fix panic when using custom EC2 API for SD #4672
[BUGFIX] Fix panic when Zookeeper SD cannot connect to servers #4669
[BUGFIX] Make the skip_head an optional parameter for snapshot API #4674
- Update to 2.4.2
+ Bug Fixes:
[BUGFIX] Handle WAL corruptions properly prometheus/tsdb#389
[BUGFIX] Handle WAL migrations correctly on Windows prometheus/tsdb#392
- Update to 2.4.1
+ New TSDB metrics
+ Bug Fixes:
Render UI correctly for Windows
- Update to 2.4.0
+ The WAL implementation has been re-written so the storage is not forward
compatible. Prometheus 2.3 storage will work on 2.4 but not vice-versa
+ Reduce remote write default retries
+ Remove /heap endpoint
+ Features:
* Persist alert 'for' state across restarts
* Add API providing per target metric metadata
* Add API providing recording and alerting rules
+ Enhancements:
* Brand new WAL implementation for TSDB. Forwards incompatible with previous WAL.
* Show rule evaluation errors in UI
* Throttle resends of alerts to Alertmanager
* Send EndsAt along with the alert to Alertmanager
* Limit the samples returned by remote read endpoint
* Limit the data read in through remote read
* Coalesce identical SD configuations
* promtool: Add new commands for debugging and querying
* Update console examples for node_exporter v0.16.0
* Optimize PromQL aggregations
* Remote read: Add Offset to hints
* consul_sd: Add support for ServiceMeta field
* ec2_sd: Maintain order of subnet_id label
* ec2_sd: Add support for custom endpoint to support EC2 compliant APIs
* ec2_sd: Add instance_owner label
* azure_sd: Add support for VMSS discovery and multiple environments
* gce_sd: Add instance_id label
* Forbid rule-abiding robots from indexing
* Log virtual memory limits on startup
+ Bug Fixes:
* Wait for service discovery to stop before exiting
* Render SD configs properly
* Only add LookbackDelta to vector selectors
* ec2_sd: Handle panic-ing nil pointer
* consul_sd: Stop leaking connections
* Use templated labels also to identify alerts
* Reduce floating point errors in stddev and related functions
* Log errors while encoding responses
- Update to 2.3.2
+ Bug Fixes:
* Fix various tsdb bugs
* Reorder startup and shutdown to prevent panics.
* Exit with non-zero code on error
* discovery/kubernetes/ingress: fix scheme discovery
* Fix race in zookeeper sd
* Better timeout handling in promql
* Propogate errors when selecting series from the tsdb
- Update to 2.3.1
+ Bug Fixes:
* Avoid infinite loop on duplicate NaN values.
* Fix nil pointer deference when using various API endpoints
* config: set target group source index during unmarshalling
* discovery/file: fix logging
* kubernetes_sd: fix namespace filtering
* web: restore old path prefix behavior
* web: remove security headers added in 2.3.0
- Update to 2.3.0
+ marathon_sd: use auth_token and auth_token_file for token-based authentication
instead of bearer_token and bearer_token_file respectively
+ Metric names for HTTP server metrics changed
+ Features:
* Add query commands to promtool
* Add security headers to HTTP server responses
* Pass query hints via remote read API
* Basic auth passwords can now be configured via file across all configuration
+ Enhancements:
* Optimise PromQL and API serialization for memory usage and allocations
* Limit number of dropped targets in web UI
* Consul and EC2 service discovery allow using server-side filtering for performance improvement
* Add advanced filtering configuration to EC2 service discovery
* marathon_sd: adds support for basic and bearer authentication, plus all
other common HTTP client options (TLS config, proxy URL, etc.)
* Provide machine type metadata and labels in GCE service discovery
* Add pod controller kind and name to Kubernetes service discovery data
* Move TSDB to flock-based log file that works with Docker containers
+ Bug Fixes:
* Properly propagate storage errors in PromQL
* Fix path prefix for web pages
* Fix goroutine leak in Consul service discovery
* Fix races in scrape manager
* Fix OOM for very large k in PromQL topk() queries
* Make remote write more resilient to unavailable receivers
* Make remote write shutdown cleanly
* Don't leak files on errors in TSDB's tombstone cleanup
* Unary minus expressions now removes the metric name from results
* Fix bug that lead to wrong amount of samples considered for time range expressions
- Update to 2.2.1
+ Bug Fixes:
* Fix data loss in TSDB on compaction
* Correctly stop timer in remote-write path
* Fix deadlock triggered by loading targets page
* Fix incorrect buffering of samples on range selection queries
* Handle large index files on windows properly
- Update to 2.2.0
+ This release introduces improvements to the storage format and fixes a
regression introduced in 2.1. As a result Prometheus servers upgraded
to 2.2 cannot be downgraded to a lower version anymore!
+ Rename file SD mtime metric
+ Send target update on empty pod IP in Kubernetes SD
+ Features:
* Add API endpoint for flags.
* Add API endpoint for dropped targets.
* Display annotations on alerts page.
* Add option to skip head data when taking snapshots
+ Enhancements:
* Federation performance improvement.
* Read bearer token file on every scrape.
* Improve typeahead on /graph page.
* Change rule file formatting.
* Set consul server default to localhost:8500.
* Add dropped Alertmanagers to API info endpoint.
* Add OS type meta label to Azure SD.
* Validate required fields in SD configuration.
+ Bug Fixes:
* Prevent stack overflow on deep recursion in TSDB.
* Correctly read offsets in index files that are greater than 4GB.
* Fix scraping behavior for empty labels.
* Drop metric name for bool modifier.
* Fix races in discovery.
* Fix Kubernetes endpoints SD for empty subsets.
* Throttle updates from SD providers, which caused increased CPU usage and allocations.
* Fix TSDB block reload issue.
* Fix PromQL printing of empty without().
* Don't reset FiredAt for inactive alerts.
* Fix erroneous file version changes and repair existing data.
- remove pr-3174.patch, has been fixed in https://github.com/prometheus/prometheus/pull/3517
- spec: fix directory for fdupes
- Update to 2.1.0:
+ Features:
* New Service Discovery UI showing labels before and after relabelling.
* New Admin APIs added to v1 to delete, snapshot and remove tombstones.
+ Enhancements:
* The graph UI autcomplete now includes your previous queries.
* Federation is now much faster for large numbers of series.
* Added new metrics to measure rule timings.
* Rule evaluation times added to the rules UI.
* Added metrics to measure modified time of file SD files.
* Kubernetes SD now includes POD UID in discovery metadata.
* The Query APIs now return optional stats on query execution times.
* The index now no longer has the 4GiB size limit and is also smaller.
+ Bug Fixes:
* Remote read read_recent option is now false by default.
* Pass the right configuration to each Alertmanager (AM) when using
multiple AM configs.
* Fix not-matchers not selecting series with labels unset.
* tsdb: Fix occasional panic in head block.
* tsdb: Close files before deletion to fix retention issues on Windows
and NFS.
* tsdb: Cleanup and do not retry failing compactions.
* tsdb: Close WAL while shutting down.
- Update to final v2.0.0 release (bsc#1067341):
+ Compared to version 1.x, this release includes a completely rewritten
storage engine, huge performance improvements, but also many backwards
incompatible changes. For more information, read the announcement blog
post and migration guide:
- https://prometheus.io/blog/2017/11/08/announcing-prometheus-2-0/
- https://prometheus.io/docs/prometheus/2.0/migration/
+ Changes:
* Completely rewritten storage layer, with WAL. This is not backwards
compatible with 1.x storage, and many flags have changed/disappeared.
* New staleness behavior. Series now marked stale after target scrapes
no longer return them, and soon after targets disappear from service
discovery.
* Rules files use YAML syntax now. Conversion tool added to promtool.
* Removed count_scalar, drop_common_labels functions and keep_common
modifier from PromQL.
* Rewritten exposition format parser with much higher performance. The
Protobuf exposition format is no longer supported.
* Example console templates updated for new storage and metrics names.
Examples other than node exporter and Prometheus removed.
* Admin and lifecycle APIs now disabled by default, can be reenabled via
flags
* Flags switched to using Kingpin, all flags are now --flagname rather
than -flagname.
+ Features:
* Remote read can be configured to not read data which is available
locally. This is enabled by default.
* Rules can be grouped now. Rules within a rule group are executed
sequentially.
* Added experimental GRPC apis
* Add timestamp() function to PromQL.
+ Enhancements:
* Remove remote read from the query path if no remote storage is
configured.
* Bump Consul HTTP client timeout to not match the Consul SD watch
timeout.
* Go-conntrack added to provide HTTP connection metrics.
+ Bug Fixes:
* Fix connection leak in Consul SD.
- Update to v2.0.0-rc.3 (bsc#1067030):
+ Enhancements:
* Remove remote read from the query path if no remote storage
is configured.
* Bump Consul HTTP client timeout to not match the Consul SD
watch timeout.
* Bump up a too small max block duration to the min block
duration instead of returning an error.
+ Bug Fixes:
* Avoid needless padding of 4 zero bytes in index files.
* Delete old blocks during reload (necessary on MS Windows).
* Fix regression of alert rules state loss on config reload.
* Serialize background WAL operations to avoid data races.
+ This also contains bugfixes and remote-storage features from
the 1.8 branch.
- Update to v2.0.0-rc.2 (bsc#1065377):
+ Enhancements:
* Handle WAL segments with corrupted header gracefully
* Stabilize memory usage during WAL replay
+ Changes:
* Prefix all storage metrics with prometheus_
+ Bug Fixes:
* Correctly handle label removal in remote read
* Fix chunk misalignment causing out-of-order samples
* Fix connection leak in Consul SD
* Handle invalid chunk dereferences gracefully
* Prevent potential deadlock during failing querier construction
- Update to v2.0.0-rc.1:
+ Features/Enhancements:
* Added a warning for time-drift between the browser and the
prometheus-server.
* Much faster WAL read-back on restart.
+ Bug Fixes:
* Fixed Remote-read to not drop the first series.
* Validate recording-rule names.
* Fix several races.
* Only close blocks if there are no iterators accessing it.
- Refresh 0002-Default-settings.patch
- Update to v2.0.0-rc.0:
+ Numerous changes to the new storage layer, the main changes being:
* Remove `count_scalar`, `keep_common` and `drop_common_labels` functions
* Breaking change in the index format for better consistency
* Fix panic due garbage collected mmap'ed strings
* Fix broken snapshots and admin APIs
* Send HTTP Accept header when scraping
* Use the WAL flush interval passed instead of the hardcoded value
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
- Refresh 0002-Default-settings.patch
- Fix loopback address for REST API gateway (bsc#1059462)
+ Added pr-3174.patch
- Update to v2.0.0-beta.5:
+ Bug Fixes:
* Remove deadlock on startup when restoring WAL
* Fix semantical races resulting in invalid persisted files
* Correctly read back WAL in certain edge cases
* Prevent crashes caused by changing metric representations in target's
/metrics
+ Enhancements:
* Overall memory usage reduction
* Serve debugging endpoints while TSDB is loading
* Healthy endpoint correctly reflects liveness during startup
* Switch to consistent usage of go-kit/log
+ This release may have issues with files written by previous beta releases
- Refresh 0002-Default-settings.patch to apply cleanly again
- Update to v2.0.0-beta.4:
+ Numerous changes to the new storage layer, the main changes being:
* Single, compacted write ahead log
* Single in-memory block with garbage collection
* Improve compression of index files
* Cache series dropped via metric_relabel_configs
* Pool byte buffers for scraping
+ Overall the changes achieve a baseline reduction in memory consumption
and reduce peak memory usage by 30-40% compared to 2.0.0-beta.2
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
- Update to v2.0.0-beta.2
+ vendor: update prometheus/tsdb (Fix panic caused by 0 division)
- Update to v2.0.0-beta.1
+ Includes a new storage layer, which reduces bottlenecks and shows
considerable performance improvements, but does not work with old v1
storage data.
- Refresh 0002-Default-settings.patch to apply cleanly again
- Add systemd unit file and default config (/etc/prometheus/prometheus.yml)
- Run as unprivileged "prometheus" user
- Default to /etc/prometheus for config and /var/lib/prometheus for metrics
+ Added 0002-Default-settings.patch
- Add BuildRequires: glibc-devel-static to fix aarch64 build
- Initial version
OBS-URL: https://build.opensuse.org/request/show/720641
OBS-URL: https://build.opensuse.org/package/show/server:monitoring/golang-github-prometheus-prometheus?expand=0&rev=17
2019-08-07 10:15:50 +02:00
+ Added 0003-Add-Uyuni-service-discovery.patch
2019-07-26 13:24:49 +02:00
2019-07-19 09:43:39 +02:00
-------------------------------------------------------------------
Thu Jul 18 01:06:13 UTC 2019 - Simon Crute <simon.crute@grafana.com>
- readded _service file removed in error.
- Update to 2.11.1
+ Bug Fix:
* Fix potential panic when prometheus is watching multiple zookeeper paths.
- Update to 2.11.0
+ Bug Fix:
* resolve race condition in maxGauge.
* Fix ZooKeeper connection leak.
* Improved atomicity of .tmp block replacement during compaction for usual case.
* Fix "unknown series references" after clean shutdown.
* Re-calculate block size when calling block.Delete.
* Fix unsafe snapshots with head block.
* prometheus_tsdb_compactions_failed_total is now incremented on any compaction failure.
+ Changes:
* Remove max_retries from queue_config (it has been unused since rewriting remote-write to utilize the write-ahead-log)
* The meta file BlockStats no longer holds size information. This is now dynamically calculated and kept in memory. It also includes the meta file size which was not included before
* Renamed metric from prometheus_tsdb_wal_reader_corruption_errors to prometheus_tsdb_wal_reader_corruption_errors_total
+ Features:
* Add option to use Alertmanager API v2.
* Added humanizePercentage function for templates.
* Include InitContainers in Kubernetes Service Discovery.
* Provide option to compress WAL records using Snappy.
+ Enhancements:
* Create new clean segment when starting the WAL.
* Reduce allocations in PromQL aggregations.
* Add storage warnings to LabelValues and LabelNames API results.
* Add prometheus_http_requests_total metric.
* Enable openbsd/arm build.
* Remote-write allocation improvements.
* Query performance improvement: Efficient iteration and search in HashForLabels and HashWithoutLabels.
* Allow injection of arbitrary headers in promtool.
* Allow passing external_labels in alert unit tests groups.
* Allows globs for rules when unit testing.
* Improved postings intersection matching.
* Reduced disk usage for WAL for small setups.
* Optimize queries using regexp for set lookups.
2019-07-06 22:18:59 +02:00
-------------------------------------------------------------------
Thu Jul 4 12:08:24 UTC 2019 - Simon Crute <simon.crute@grafana.com>
- rebase patch002-Default-settings.patch
- Update to 2.10.0:
+ Bug Fixes:
* TSDB: Don't panic when running out of disk space and recover nicely from the condition
* TSDB: Correctly handle empty labels.
* TSDB: Don't crash on an unknown tombstone reference.
* Storage/remote: Remove queue-manager specific metrics if queue no longer exists.
* PromQL: Correctly display {__name__="a"}.
* Discovery/kubernetes: Use service rather than ingress as the name for the service workqueue.
* Discovery/azure: Don't panic on a VM with a public IP.
* Web: Fixed Content-Type for js and css instead of using /etc/mime.types.
* API: Encode alert values as string to correctly represent Inf/NaN.
+ Features:
* Template expansion: Make external labels available as $externalLabels in alert and console template expansion.
* TSDB: Add prometheus_tsdb_wal_segment_current metric for the WAL segment index that TSDB is currently writing to. tsdb
* Scrape: Add scrape_series_added per-scrape metric. #5546
+ Enhancements
* Discovery/kubernetes: Add labels __meta_kubernetes_endpoint_node_name and __meta_kubernetes_endpoint_hostname.
* Discovery/azure: Add label __meta_azure_machine_public_ip.
* TSDB: Simplify mergedPostings.Seek, resulting in better performance if there are many posting lists. tsdb
* Log filesystem type on startup.
* Cmd/promtool: Use POST requests for Query and QueryRange. client_golang
* Web: Sort alerts by group name.
* Console templates: Add convenience variables $rawParams, $params, $path.
- Upadte to 2.9.2
+ Bug Fixes:
* Make sure subquery range is taken into account for selection
* Exhaust every request body before closing it
* Cmd/promtool: return errors from rule evaluations
* Remote Storage: string interner should not panic in release
* Fix memory allocation regression in mergedPostings.Seek tsdb
- Update to 2.9.1
+ Bug Fixes:
* Discovery/kubernetes: fix missing label sanitization
* Remote_write: Prevent reshard concurrent with calling stop
- Update to 2.9.0
+ Feature:
* Add honor_timestamps scrape option.
+ Enhancements:
* Update Consul to support catalog.ServiceMultipleTags.
* Discovery/kubernetes: add present labels for labels/annotations.
* OpenStack SD: Add ProjectID and UserID meta labels.
* Add GODEBUG and retention to the runtime page.
* Add support for POSTing to /series endpoint.
* Support PUT methods for Lifecycle and Admin APIs.
* Scrape: Add global jitter for HA server.
* Check for cancellation on every step of a range evaluation.
* String interning for labels & values in the remote_write path.
* Don't lose the scrape cache on a failed scrape.
* Reload cert files from disk automatically. common
* Use fixed length millisecond timestamp format for logs. common
* Performance improvements for postings.
Bug Fixes:
* Remote Write: fix checkpoint reading.
* Check if label value is valid when unmarshaling external labels from YAML.
* Promparse: sort all labels when parsing.
* Reload rules: copy state on both name and labels.
* Exponentation operator to drop metric name in result of operation.
* Config: resolve more file paths.
* Promtool: resolve relative paths in alert test files.
* Set TLSHandshakeTimeout in HTTP transport. common
* Use fsync to be more resilient to machine crashes.
* Keep series that are still in WAL in checkpoints.
- Update to 2.8.1
+ Bug Fixes
* Display the job labels in /targets which was removed accidentally
- Update to 2.8.0
+ Change:
* This release uses Write-Ahead Logging (WAL) for the remote_write API. This currently causes a slight increase in memory usage, which will be addressed in future releases.
* Default time retention is used only when no size based retention is specified. These are flags where time retention is specified by the flag --storage.tsdb.retention and size retention by --storage.tsdb.retention.size.
* prometheus_tsdb_storage_blocks_bytes_total is now prometheus_tsdb_storage_blocks_bytes.
+ Feature:
* (EXPERIMENTAL) Time overlapping blocks are now allowed; vertical compaction and vertical query merge. It is an optional feature which is controlled by the --storage.tsdb.allow-overlapping-blocks flag, disabled by default.
+ Enhancements:
* Use the WAL for remote_write API.
* Query performance improvements.
* UI enhancements with upgrade to Bootstrap 4.
* Reduce time that Alertmanagers are in flux when reloaded.
* Limit number of metrics displayed on UI to 10000.
* (1) Remember All/Unhealthy choice on target-overview when reloading page. (2) Resize text-input area on Graph page on mouseclick.
* In histogram_quantile merge buckets with equivalent le values.
* Show list of offending labels in the error message in many-to-many scenarios.
* Show Storage Retention criteria in effect on /status page.
+ Bug Fixes:
+ Fix sorting of rule groups.
+ Fix support for password_file and bearer_token_file in Kubernetes SD.
+ Scrape: catch errors when creating HTTP clients
+ Adds new metrics:
prometheus_target_scrape_pools_total
prometheus_target_scrape_pools_failed_total
prometheus_target_scrape_pool_reloads_total
prometheus_target_scrape_pool_reloads_failed_total
+ Fix panic when aggregator param is not a literal.
2019-02-22 10:44:31 +01:00
-------------------------------------------------------------------
Fri Feb 22 09:20:11 UTC 2019 - Jan Fajerski <jan.fajerski@suse.com>
- fix spec file: actually ship promtool
2018-03-14 09:37:06 +01:00
-------------------------------------------------------------------
Accepting request 672738 from home:jfajerski:branches:server:monitoring
- Update to 2.7.1:
+ Bug Fixes:
* Fix a Stored DOM XSS vulnerability with query history (boo#1124610)
* prometheus_rule_group_last_duration_seconds now reports seconds instead of nanoseconds
* Make sure the targets are consistently sorted in the targets page
- Update to 2.7.0:
+ cli flag depreacted: storage.tsdb.retention use storage.tsdb.retention.time
instead; depreacted flag will be removed in 3.0
+ Features:
* Add subqueries to PromQL
* Add support for disk size based retention. Note that we don't
consider the WAL size which could be significant and the time
based retention policy also applies (experimental)
* Add CORS origin flag
+ Bug Fixes:
* Don't depend on given order when comparing samples in alert unit testing
* Make sure the retention period doesn't overflow
* Don't generate blocks with no samples
- Update to 2.6.0:
+ Remove default flags from the container's entrypoint, run Prometheus from
/etc/prometheus and symlink the storage directory to /etc/prometheus/data
+ Promtool: Remove the update command
+ Features:
* Add JSON log format via the --log.format flag
* API: Add /api/v1/labels endpoint to get all label names
* Web: Allow setting the page's title via the --web.ui-title flag
+ Enhancements:
* Add prometheus_tsdb_lowest_timestamp_seconds, prometheus_tsdb_head_min_time_seconds
and prometheus_tsdb_head_max_time_seconds metrics
* Add rule_group_last_evaluation_timestamp_seconds metric
* Add prometheus_template_text_expansion_failures_total and prometheus_template_text_expansions_total metrics
* Set consistent User-Agent header in outgoing requests
* Azure SD: Error out at load time when authentication parameters are missing
* EC2 SD: Add the machine's private DNS name to the discovery metadata
* EC2 SD: Add the operating system's platform to the discovery metadata
* Kubernetes SD: Add the pod's phase to the discovery metadata
* Kubernetes SD: Log Kubernetes messages
* Promtool: Collect CPU and trace profiles
* Promtool: Support writing output as JSON
* Remote Read: Return available data if remote read fails partially
* Remote Write: Improve queue performance
* Remote Write: Add min_shards parameter to set the minimum number of shards
* TSDB: Improve WAL reading
* TSDB: Memory improvements
* Web: Log stack traces on panic
* Web UI: Add copy to clipboard button for configuration
* Web UI: Support console queries at specific times
* Web UI: group targets by job then instance
+ Bug Fixes:
* Deduplicate handler labels for HTTP metrics
* Fix leaked queriers causing shutdowns to hang
* Fix configuration loading panics on nil pointer slice elements
* API: Correctly skip mismatching targets on /api/v1/targets/metadata
* API: Better rounding for incoming query timestamps
* Discovery: Remove all targets when the scrape configuration gets empty
* PromQL: Fix a goroutine leak in the lexer/parser
* Scrape: Fix deadlock in the scrape's manager
* Scrape: Scrape targets at fixed intervals even after Prometheus restarts
* TSDB: Support restored snapshots including the head properly
* TSDB: Repair WAL when the last record in a segment is torn
- Update to 2.5.0
+ Group targets by scrape config instead of job name
+ Marathon SD: Various changes to adapt to Marathon 1.5+
+ Discovery: Split prometheus_sd_discovered_targets metric by scrape and
notify (Alertmanager SD) as well as by section in the respective configuration
+ Enhancements:
* Support s390x platform for Linux
* API: Add prometheus_api_remote_read_queries metric tracking currently
executed or waiting remote read API requests
* Remote Read: Add prometheus_remote_storage_remote_read_queries metric
tracking currently in-flight remote read queries
* Remote Read: Reduced memory usage
* Discovery: Add prometheus_sd_discovered_targets,
prometheus_sd_received_updates_total, prometheus_sd_updates_delayed_total,
and prometheus_sd_updates_total metrics for discovery subsystem
* Discovery: Improve performance of previously slow updates of changes of targets
* Kubernetes SD: Add extended metrics
* OpenStack SD: Support discovering instances from all projects
* OpenStack SD: Discover all interfaces
* OpenStack SD: Support tls_config for the used HTTP client
* Triton SD: Add ability to filter triton_sd targets by pre-defined groups
* Web UI: Avoid browser spell-checking in expression field
* Web UI: Add scrape duration and last evaluation time in targets and rules pages
* Web UI: Improve rule view by wrapping lines
* Rules: Error out at load time for invalid templates, rather than at evaluation time
+ Bug Fixes:
* Change max/min over_time to handle NaNs properly
* Check label name for count_values PromQL function
* Ensure that vectors and matrices do not contain identical label-sets
- Update to 2.4.3
+ Bug Fixes:
[BUGFIX] Fix panic when using custom EC2 API for SD #4672
[BUGFIX] Fix panic when Zookeeper SD cannot connect to servers #4669
[BUGFIX] Make the skip_head an optional parameter for snapshot API #4674
- Update to 2.4.2
+ Bug Fixes:
[BUGFIX] Handle WAL corruptions properly prometheus/tsdb#389
[BUGFIX] Handle WAL migrations correctly on Windows prometheus/tsdb#392
- Update to 2.4.1
+ New TSDB metrics
+ Bug Fixes:
Render UI correctly for Windows
- Update to 2.4.0
+ The WAL implementation has been re-written so the storage is not forward
compatible. Prometheus 2.3 storage will work on 2.4 but not vice-versa
+ Reduce remote write default retries
+ Remove /heap endpoint
+ Features:
* Persist alert 'for' state across restarts
* Add API providing per target metric metadata
* Add API providing recording and alerting rules
+ Enhancements:
* Brand new WAL implementation for TSDB. Forwards incompatible with previous WAL.
* Show rule evaluation errors in UI
* Throttle resends of alerts to Alertmanager
* Send EndsAt along with the alert to Alertmanager
* Limit the samples returned by remote read endpoint
* Limit the data read in through remote read
* Coalesce identical SD configuations
* promtool: Add new commands for debugging and querying
* Update console examples for node_exporter v0.16.0
* Optimize PromQL aggregations
* Remote read: Add Offset to hints
* consul_sd: Add support for ServiceMeta field
* ec2_sd: Maintain order of subnet_id label
* ec2_sd: Add support for custom endpoint to support EC2 compliant APIs
* ec2_sd: Add instance_owner label
* azure_sd: Add support for VMSS discovery and multiple environments
* gce_sd: Add instance_id label
* Forbid rule-abiding robots from indexing
* Log virtual memory limits on startup
+ Bug Fixes:
* Wait for service discovery to stop before exiting
* Render SD configs properly
* Only add LookbackDelta to vector selectors
* ec2_sd: Handle panic-ing nil pointer
* consul_sd: Stop leaking connections
* Use templated labels also to identify alerts
* Reduce floating point errors in stddev and related functions
* Log errors while encoding responses
- Update to 2.3.2
+ Bug Fixes:
* Fix various tsdb bugs
* Reorder startup and shutdown to prevent panics.
* Exit with non-zero code on error
* discovery/kubernetes/ingress: fix scheme discovery
* Fix race in zookeeper sd
* Better timeout handling in promql
* Propogate errors when selecting series from the tsdb
- Update to 2.3.1
+ Bug Fixes:
* Avoid infinite loop on duplicate NaN values.
* Fix nil pointer deference when using various API endpoints
* config: set target group source index during unmarshalling
* discovery/file: fix logging
* kubernetes_sd: fix namespace filtering
* web: restore old path prefix behavior
* web: remove security headers added in 2.3.0
- Update to 2.3.0
+ marathon_sd: use auth_token and auth_token_file for token-based authentication
instead of bearer_token and bearer_token_file respectively
+ Metric names for HTTP server metrics changed
+ Features:
* Add query commands to promtool
* Add security headers to HTTP server responses
* Pass query hints via remote read API
* Basic auth passwords can now be configured via file across all configuration
+ Enhancements:
* Optimise PromQL and API serialization for memory usage and allocations
* Limit number of dropped targets in web UI
* Consul and EC2 service discovery allow using server-side filtering for performance improvement
* Add advanced filtering configuration to EC2 service discovery
* marathon_sd: adds support for basic and bearer authentication, plus all
other common HTTP client options (TLS config, proxy URL, etc.)
* Provide machine type metadata and labels in GCE service discovery
* Add pod controller kind and name to Kubernetes service discovery data
* Move TSDB to flock-based log file that works with Docker containers
+ Bug Fixes:
* Properly propagate storage errors in PromQL
* Fix path prefix for web pages
* Fix goroutine leak in Consul service discovery
* Fix races in scrape manager
* Fix OOM for very large k in PromQL topk() queries
* Make remote write more resilient to unavailable receivers
* Make remote write shutdown cleanly
* Don't leak files on errors in TSDB's tombstone cleanup
* Unary minus expressions now removes the metric name from results
* Fix bug that lead to wrong amount of samples considered for time range expressions
- Update to 2.2.1
+ Bug Fixes:
* Fix data loss in TSDB on compaction
* Correctly stop timer in remote-write path
* Fix deadlock triggered by loading targets page
* Fix incorrect buffering of samples on range selection queries
* Handle large index files on windows properly
- Update to 2.2.0
+ This release introduces improvements to the storage format and fixes a
regression introduced in 2.1. As a result Prometheus servers upgraded
to 2.2 cannot be downgraded to a lower version anymore!
+ Rename file SD mtime metric
+ Send target update on empty pod IP in Kubernetes SD
+ Features:
* Add API endpoint for flags.
* Add API endpoint for dropped targets.
* Display annotations on alerts page.
* Add option to skip head data when taking snapshots
+ Enhancements:
* Federation performance improvement.
* Read bearer token file on every scrape.
* Improve typeahead on /graph page.
* Change rule file formatting.
* Set consul server default to localhost:8500.
* Add dropped Alertmanagers to API info endpoint.
* Add OS type meta label to Azure SD.
* Validate required fields in SD configuration.
+ Bug Fixes:
* Prevent stack overflow on deep recursion in TSDB.
* Correctly read offsets in index files that are greater than 4GB.
* Fix scraping behavior for empty labels.
* Drop metric name for bool modifier.
* Fix races in discovery.
* Fix Kubernetes endpoints SD for empty subsets.
* Throttle updates from SD providers, which caused increased CPU usage and allocations.
* Fix TSDB block reload issue.
* Fix PromQL printing of empty without().
* Don't reset FiredAt for inactive alerts.
* Fix erroneous file version changes and repair existing data.
OBS-URL: https://build.opensuse.org/request/show/672738
OBS-URL: https://build.opensuse.org/package/show/server:monitoring/golang-github-prometheus-prometheus?expand=0&rev=7
2019-02-08 11:30:26 +01:00
Fri Feb 8 07:24:59 UTC 2019 - Jan Fajerski <jan.fajerski@suse.com>
- Update to 2.7.1:
+ Bug Fixes:
* Fix a Stored DOM XSS vulnerability with query history (boo#1124610)
* prometheus_rule_group_last_duration_seconds now reports seconds instead of nanoseconds
* Make sure the targets are consistently sorted in the targets page
- Update to 2.7.0:
+ cli flag depreacted: storage.tsdb.retention use storage.tsdb.retention.time
instead; depreacted flag will be removed in 3.0
+ Features:
* Add subqueries to PromQL
* Add support for disk size based retention. Note that we don't
consider the WAL size which could be significant and the time
based retention policy also applies (experimental)
* Add CORS origin flag
+ Bug Fixes:
* Don't depend on given order when comparing samples in alert unit testing
* Make sure the retention period doesn't overflow
* Don't generate blocks with no samples
- Update to 2.6.0:
+ Remove default flags from the container's entrypoint, run Prometheus from
/etc/prometheus and symlink the storage directory to /etc/prometheus/data
+ Promtool: Remove the update command
+ Features:
* Add JSON log format via the --log.format flag
* API: Add /api/v1/labels endpoint to get all label names
* Web: Allow setting the page's title via the --web.ui-title flag
+ Enhancements:
* Add prometheus_tsdb_lowest_timestamp_seconds, prometheus_tsdb_head_min_time_seconds
and prometheus_tsdb_head_max_time_seconds metrics
* Add rule_group_last_evaluation_timestamp_seconds metric
* Add prometheus_template_text_expansion_failures_total and prometheus_template_text_expansions_total metrics
* Set consistent User-Agent header in outgoing requests
* Azure SD: Error out at load time when authentication parameters are missing
* EC2 SD: Add the machine's private DNS name to the discovery metadata
* EC2 SD: Add the operating system's platform to the discovery metadata
* Kubernetes SD: Add the pod's phase to the discovery metadata
* Kubernetes SD: Log Kubernetes messages
* Promtool: Collect CPU and trace profiles
* Promtool: Support writing output as JSON
* Remote Read: Return available data if remote read fails partially
* Remote Write: Improve queue performance
* Remote Write: Add min_shards parameter to set the minimum number of shards
* TSDB: Improve WAL reading
* TSDB: Memory improvements
* Web: Log stack traces on panic
* Web UI: Add copy to clipboard button for configuration
* Web UI: Support console queries at specific times
* Web UI: group targets by job then instance
+ Bug Fixes:
* Deduplicate handler labels for HTTP metrics
* Fix leaked queriers causing shutdowns to hang
* Fix configuration loading panics on nil pointer slice elements
* API: Correctly skip mismatching targets on /api/v1/targets/metadata
* API: Better rounding for incoming query timestamps
* Discovery: Remove all targets when the scrape configuration gets empty
* PromQL: Fix a goroutine leak in the lexer/parser
* Scrape: Fix deadlock in the scrape's manager
* Scrape: Scrape targets at fixed intervals even after Prometheus restarts
* TSDB: Support restored snapshots including the head properly
* TSDB: Repair WAL when the last record in a segment is torn
- Update to 2.5.0
+ Group targets by scrape config instead of job name
+ Marathon SD: Various changes to adapt to Marathon 1.5+
+ Discovery: Split prometheus_sd_discovered_targets metric by scrape and
notify (Alertmanager SD) as well as by section in the respective configuration
+ Enhancements:
* Support s390x platform for Linux
* API: Add prometheus_api_remote_read_queries metric tracking currently
executed or waiting remote read API requests
* Remote Read: Add prometheus_remote_storage_remote_read_queries metric
tracking currently in-flight remote read queries
* Remote Read: Reduced memory usage
* Discovery: Add prometheus_sd_discovered_targets,
prometheus_sd_received_updates_total, prometheus_sd_updates_delayed_total,
and prometheus_sd_updates_total metrics for discovery subsystem
* Discovery: Improve performance of previously slow updates of changes of targets
* Kubernetes SD: Add extended metrics
* OpenStack SD: Support discovering instances from all projects
* OpenStack SD: Discover all interfaces
* OpenStack SD: Support tls_config for the used HTTP client
* Triton SD: Add ability to filter triton_sd targets by pre-defined groups
* Web UI: Avoid browser spell-checking in expression field
* Web UI: Add scrape duration and last evaluation time in targets and rules pages
* Web UI: Improve rule view by wrapping lines
* Rules: Error out at load time for invalid templates, rather than at evaluation time
+ Bug Fixes:
* Change max/min over_time to handle NaNs properly
* Check label name for count_values PromQL function
* Ensure that vectors and matrices do not contain identical label-sets
- Update to 2.4.3
+ Bug Fixes:
[BUGFIX] Fix panic when using custom EC2 API for SD #4672
[BUGFIX] Fix panic when Zookeeper SD cannot connect to servers #4669
[BUGFIX] Make the skip_head an optional parameter for snapshot API #4674
- Update to 2.4.2
+ Bug Fixes:
[BUGFIX] Handle WAL corruptions properly prometheus/tsdb#389
[BUGFIX] Handle WAL migrations correctly on Windows prometheus/tsdb#392
- Update to 2.4.1
+ New TSDB metrics
+ Bug Fixes:
Render UI correctly for Windows
- Update to 2.4.0
+ The WAL implementation has been re-written so the storage is not forward
compatible. Prometheus 2.3 storage will work on 2.4 but not vice-versa
+ Reduce remote write default retries
+ Remove /heap endpoint
+ Features:
* Persist alert 'for' state across restarts
* Add API providing per target metric metadata
* Add API providing recording and alerting rules
+ Enhancements:
* Brand new WAL implementation for TSDB. Forwards incompatible with previous WAL.
* Show rule evaluation errors in UI
* Throttle resends of alerts to Alertmanager
* Send EndsAt along with the alert to Alertmanager
* Limit the samples returned by remote read endpoint
* Limit the data read in through remote read
* Coalesce identical SD configuations
* promtool: Add new commands for debugging and querying
* Update console examples for node_exporter v0.16.0
* Optimize PromQL aggregations
* Remote read: Add Offset to hints
* consul_sd: Add support for ServiceMeta field
* ec2_sd: Maintain order of subnet_id label
* ec2_sd: Add support for custom endpoint to support EC2 compliant APIs
* ec2_sd: Add instance_owner label
* azure_sd: Add support for VMSS discovery and multiple environments
* gce_sd: Add instance_id label
* Forbid rule-abiding robots from indexing
* Log virtual memory limits on startup
+ Bug Fixes:
* Wait for service discovery to stop before exiting
* Render SD configs properly
* Only add LookbackDelta to vector selectors
* ec2_sd: Handle panic-ing nil pointer
* consul_sd: Stop leaking connections
* Use templated labels also to identify alerts
* Reduce floating point errors in stddev and related functions
* Log errors while encoding responses
- Update to 2.3.2
+ Bug Fixes:
* Fix various tsdb bugs
* Reorder startup and shutdown to prevent panics.
* Exit with non-zero code on error
* discovery/kubernetes/ingress: fix scheme discovery
* Fix race in zookeeper sd
* Better timeout handling in promql
* Propogate errors when selecting series from the tsdb
- Update to 2.3.1
+ Bug Fixes:
* Avoid infinite loop on duplicate NaN values.
* Fix nil pointer deference when using various API endpoints
* config: set target group source index during unmarshalling
* discovery/file: fix logging
* kubernetes_sd: fix namespace filtering
* web: restore old path prefix behavior
* web: remove security headers added in 2.3.0
- Update to 2.3.0
+ marathon_sd: use auth_token and auth_token_file for token-based authentication
instead of bearer_token and bearer_token_file respectively
+ Metric names for HTTP server metrics changed
+ Features:
* Add query commands to promtool
* Add security headers to HTTP server responses
* Pass query hints via remote read API
* Basic auth passwords can now be configured via file across all configuration
+ Enhancements:
* Optimise PromQL and API serialization for memory usage and allocations
* Limit number of dropped targets in web UI
* Consul and EC2 service discovery allow using server-side filtering for performance improvement
* Add advanced filtering configuration to EC2 service discovery
* marathon_sd: adds support for basic and bearer authentication, plus all
other common HTTP client options (TLS config, proxy URL, etc.)
* Provide machine type metadata and labels in GCE service discovery
* Add pod controller kind and name to Kubernetes service discovery data
* Move TSDB to flock-based log file that works with Docker containers
+ Bug Fixes:
* Properly propagate storage errors in PromQL
* Fix path prefix for web pages
* Fix goroutine leak in Consul service discovery
* Fix races in scrape manager
* Fix OOM for very large k in PromQL topk() queries
* Make remote write more resilient to unavailable receivers
* Make remote write shutdown cleanly
* Don't leak files on errors in TSDB's tombstone cleanup
* Unary minus expressions now removes the metric name from results
* Fix bug that lead to wrong amount of samples considered for time range expressions
- Update to 2.2.1
+ Bug Fixes:
* Fix data loss in TSDB on compaction
* Correctly stop timer in remote-write path
* Fix deadlock triggered by loading targets page
* Fix incorrect buffering of samples on range selection queries
* Handle large index files on windows properly
- Update to 2.2.0
+ This release introduces improvements to the storage format and fixes a
regression introduced in 2.1. As a result Prometheus servers upgraded
to 2.2 cannot be downgraded to a lower version anymore!
+ Rename file SD mtime metric
+ Send target update on empty pod IP in Kubernetes SD
+ Features:
* Add API endpoint for flags.
* Add API endpoint for dropped targets.
* Display annotations on alerts page.
* Add option to skip head data when taking snapshots
+ Enhancements:
* Federation performance improvement.
* Read bearer token file on every scrape.
* Improve typeahead on /graph page.
* Change rule file formatting.
* Set consul server default to localhost:8500.
* Add dropped Alertmanagers to API info endpoint.
* Add OS type meta label to Azure SD.
* Validate required fields in SD configuration.
+ Bug Fixes:
* Prevent stack overflow on deep recursion in TSDB.
* Correctly read offsets in index files that are greater than 4GB.
* Fix scraping behavior for empty labels.
* Drop metric name for bool modifier.
* Fix races in discovery.
* Fix Kubernetes endpoints SD for empty subsets.
* Throttle updates from SD providers, which caused increased CPU usage and allocations.
* Fix TSDB block reload issue.
* Fix PromQL printing of empty without().
* Don't reset FiredAt for inactive alerts.
* Fix erroneous file version changes and repair existing data.
2019-02-11 12:32:54 +01:00
- remove pr-3174.patch, has been fixed in https://github.com/prometheus/prometheus/pull/3517
Accepting request 672738 from home:jfajerski:branches:server:monitoring
- Update to 2.7.1:
+ Bug Fixes:
* Fix a Stored DOM XSS vulnerability with query history (boo#1124610)
* prometheus_rule_group_last_duration_seconds now reports seconds instead of nanoseconds
* Make sure the targets are consistently sorted in the targets page
- Update to 2.7.0:
+ cli flag depreacted: storage.tsdb.retention use storage.tsdb.retention.time
instead; depreacted flag will be removed in 3.0
+ Features:
* Add subqueries to PromQL
* Add support for disk size based retention. Note that we don't
consider the WAL size which could be significant and the time
based retention policy also applies (experimental)
* Add CORS origin flag
+ Bug Fixes:
* Don't depend on given order when comparing samples in alert unit testing
* Make sure the retention period doesn't overflow
* Don't generate blocks with no samples
- Update to 2.6.0:
+ Remove default flags from the container's entrypoint, run Prometheus from
/etc/prometheus and symlink the storage directory to /etc/prometheus/data
+ Promtool: Remove the update command
+ Features:
* Add JSON log format via the --log.format flag
* API: Add /api/v1/labels endpoint to get all label names
* Web: Allow setting the page's title via the --web.ui-title flag
+ Enhancements:
* Add prometheus_tsdb_lowest_timestamp_seconds, prometheus_tsdb_head_min_time_seconds
and prometheus_tsdb_head_max_time_seconds metrics
* Add rule_group_last_evaluation_timestamp_seconds metric
* Add prometheus_template_text_expansion_failures_total and prometheus_template_text_expansions_total metrics
* Set consistent User-Agent header in outgoing requests
* Azure SD: Error out at load time when authentication parameters are missing
* EC2 SD: Add the machine's private DNS name to the discovery metadata
* EC2 SD: Add the operating system's platform to the discovery metadata
* Kubernetes SD: Add the pod's phase to the discovery metadata
* Kubernetes SD: Log Kubernetes messages
* Promtool: Collect CPU and trace profiles
* Promtool: Support writing output as JSON
* Remote Read: Return available data if remote read fails partially
* Remote Write: Improve queue performance
* Remote Write: Add min_shards parameter to set the minimum number of shards
* TSDB: Improve WAL reading
* TSDB: Memory improvements
* Web: Log stack traces on panic
* Web UI: Add copy to clipboard button for configuration
* Web UI: Support console queries at specific times
* Web UI: group targets by job then instance
+ Bug Fixes:
* Deduplicate handler labels for HTTP metrics
* Fix leaked queriers causing shutdowns to hang
* Fix configuration loading panics on nil pointer slice elements
* API: Correctly skip mismatching targets on /api/v1/targets/metadata
* API: Better rounding for incoming query timestamps
* Discovery: Remove all targets when the scrape configuration gets empty
* PromQL: Fix a goroutine leak in the lexer/parser
* Scrape: Fix deadlock in the scrape's manager
* Scrape: Scrape targets at fixed intervals even after Prometheus restarts
* TSDB: Support restored snapshots including the head properly
* TSDB: Repair WAL when the last record in a segment is torn
- Update to 2.5.0
+ Group targets by scrape config instead of job name
+ Marathon SD: Various changes to adapt to Marathon 1.5+
+ Discovery: Split prometheus_sd_discovered_targets metric by scrape and
notify (Alertmanager SD) as well as by section in the respective configuration
+ Enhancements:
* Support s390x platform for Linux
* API: Add prometheus_api_remote_read_queries metric tracking currently
executed or waiting remote read API requests
* Remote Read: Add prometheus_remote_storage_remote_read_queries metric
tracking currently in-flight remote read queries
* Remote Read: Reduced memory usage
* Discovery: Add prometheus_sd_discovered_targets,
prometheus_sd_received_updates_total, prometheus_sd_updates_delayed_total,
and prometheus_sd_updates_total metrics for discovery subsystem
* Discovery: Improve performance of previously slow updates of changes of targets
* Kubernetes SD: Add extended metrics
* OpenStack SD: Support discovering instances from all projects
* OpenStack SD: Discover all interfaces
* OpenStack SD: Support tls_config for the used HTTP client
* Triton SD: Add ability to filter triton_sd targets by pre-defined groups
* Web UI: Avoid browser spell-checking in expression field
* Web UI: Add scrape duration and last evaluation time in targets and rules pages
* Web UI: Improve rule view by wrapping lines
* Rules: Error out at load time for invalid templates, rather than at evaluation time
+ Bug Fixes:
* Change max/min over_time to handle NaNs properly
* Check label name for count_values PromQL function
* Ensure that vectors and matrices do not contain identical label-sets
- Update to 2.4.3
+ Bug Fixes:
[BUGFIX] Fix panic when using custom EC2 API for SD #4672
[BUGFIX] Fix panic when Zookeeper SD cannot connect to servers #4669
[BUGFIX] Make the skip_head an optional parameter for snapshot API #4674
- Update to 2.4.2
+ Bug Fixes:
[BUGFIX] Handle WAL corruptions properly prometheus/tsdb#389
[BUGFIX] Handle WAL migrations correctly on Windows prometheus/tsdb#392
- Update to 2.4.1
+ New TSDB metrics
+ Bug Fixes:
Render UI correctly for Windows
- Update to 2.4.0
+ The WAL implementation has been re-written so the storage is not forward
compatible. Prometheus 2.3 storage will work on 2.4 but not vice-versa
+ Reduce remote write default retries
+ Remove /heap endpoint
+ Features:
* Persist alert 'for' state across restarts
* Add API providing per target metric metadata
* Add API providing recording and alerting rules
+ Enhancements:
* Brand new WAL implementation for TSDB. Forwards incompatible with previous WAL.
* Show rule evaluation errors in UI
* Throttle resends of alerts to Alertmanager
* Send EndsAt along with the alert to Alertmanager
* Limit the samples returned by remote read endpoint
* Limit the data read in through remote read
* Coalesce identical SD configuations
* promtool: Add new commands for debugging and querying
* Update console examples for node_exporter v0.16.0
* Optimize PromQL aggregations
* Remote read: Add Offset to hints
* consul_sd: Add support for ServiceMeta field
* ec2_sd: Maintain order of subnet_id label
* ec2_sd: Add support for custom endpoint to support EC2 compliant APIs
* ec2_sd: Add instance_owner label
* azure_sd: Add support for VMSS discovery and multiple environments
* gce_sd: Add instance_id label
* Forbid rule-abiding robots from indexing
* Log virtual memory limits on startup
+ Bug Fixes:
* Wait for service discovery to stop before exiting
* Render SD configs properly
* Only add LookbackDelta to vector selectors
* ec2_sd: Handle panic-ing nil pointer
* consul_sd: Stop leaking connections
* Use templated labels also to identify alerts
* Reduce floating point errors in stddev and related functions
* Log errors while encoding responses
- Update to 2.3.2
+ Bug Fixes:
* Fix various tsdb bugs
* Reorder startup and shutdown to prevent panics.
* Exit with non-zero code on error
* discovery/kubernetes/ingress: fix scheme discovery
* Fix race in zookeeper sd
* Better timeout handling in promql
* Propogate errors when selecting series from the tsdb
- Update to 2.3.1
+ Bug Fixes:
* Avoid infinite loop on duplicate NaN values.
* Fix nil pointer deference when using various API endpoints
* config: set target group source index during unmarshalling
* discovery/file: fix logging
* kubernetes_sd: fix namespace filtering
* web: restore old path prefix behavior
* web: remove security headers added in 2.3.0
- Update to 2.3.0
+ marathon_sd: use auth_token and auth_token_file for token-based authentication
instead of bearer_token and bearer_token_file respectively
+ Metric names for HTTP server metrics changed
+ Features:
* Add query commands to promtool
* Add security headers to HTTP server responses
* Pass query hints via remote read API
* Basic auth passwords can now be configured via file across all configuration
+ Enhancements:
* Optimise PromQL and API serialization for memory usage and allocations
* Limit number of dropped targets in web UI
* Consul and EC2 service discovery allow using server-side filtering for performance improvement
* Add advanced filtering configuration to EC2 service discovery
* marathon_sd: adds support for basic and bearer authentication, plus all
other common HTTP client options (TLS config, proxy URL, etc.)
* Provide machine type metadata and labels in GCE service discovery
* Add pod controller kind and name to Kubernetes service discovery data
* Move TSDB to flock-based log file that works with Docker containers
+ Bug Fixes:
* Properly propagate storage errors in PromQL
* Fix path prefix for web pages
* Fix goroutine leak in Consul service discovery
* Fix races in scrape manager
* Fix OOM for very large k in PromQL topk() queries
* Make remote write more resilient to unavailable receivers
* Make remote write shutdown cleanly
* Don't leak files on errors in TSDB's tombstone cleanup
* Unary minus expressions now removes the metric name from results
* Fix bug that lead to wrong amount of samples considered for time range expressions
- Update to 2.2.1
+ Bug Fixes:
* Fix data loss in TSDB on compaction
* Correctly stop timer in remote-write path
* Fix deadlock triggered by loading targets page
* Fix incorrect buffering of samples on range selection queries
* Handle large index files on windows properly
- Update to 2.2.0
+ This release introduces improvements to the storage format and fixes a
regression introduced in 2.1. As a result Prometheus servers upgraded
to 2.2 cannot be downgraded to a lower version anymore!
+ Rename file SD mtime metric
+ Send target update on empty pod IP in Kubernetes SD
+ Features:
* Add API endpoint for flags.
* Add API endpoint for dropped targets.
* Display annotations on alerts page.
* Add option to skip head data when taking snapshots
+ Enhancements:
* Federation performance improvement.
* Read bearer token file on every scrape.
* Improve typeahead on /graph page.
* Change rule file formatting.
* Set consul server default to localhost:8500.
* Add dropped Alertmanagers to API info endpoint.
* Add OS type meta label to Azure SD.
* Validate required fields in SD configuration.
+ Bug Fixes:
* Prevent stack overflow on deep recursion in TSDB.
* Correctly read offsets in index files that are greater than 4GB.
* Fix scraping behavior for empty labels.
* Drop metric name for bool modifier.
* Fix races in discovery.
* Fix Kubernetes endpoints SD for empty subsets.
* Throttle updates from SD providers, which caused increased CPU usage and allocations.
* Fix TSDB block reload issue.
* Fix PromQL printing of empty without().
* Don't reset FiredAt for inactive alerts.
* Fix erroneous file version changes and repair existing data.
OBS-URL: https://build.opensuse.org/request/show/672738
OBS-URL: https://build.opensuse.org/package/show/server:monitoring/golang-github-prometheus-prometheus?expand=0&rev=7
2019-02-08 11:30:26 +01:00
-------------------------------------------------------------------
2018-03-14 09:37:06 +01:00
Wed Mar 14 07:20:23 UTC 2018 - jan.fajerski@suse.com
- spec: fix directory for fdupes
2018-01-29 13:36:06 +01:00
-------------------------------------------------------------------
Thu Jan 25 10:11:26 UTC 2018 - kkaempf@suse.com
2018-03-03 09:28:17 +01:00
- Update to 2.1.0:
+ Features:
* New Service Discovery UI showing labels before and after relabelling.
* New Admin APIs added to v1 to delete, snapshot and remove tombstones.
+ Enhancements:
* The graph UI autcomplete now includes your previous queries.
* Federation is now much faster for large numbers of series.
* Added new metrics to measure rule timings.
* Rule evaluation times added to the rules UI.
* Added metrics to measure modified time of file SD files.
* Kubernetes SD now includes POD UID in discovery metadata.
* The Query APIs now return optional stats on query execution times.
* The index now no longer has the 4GiB size limit and is also smaller.
+ Bug Fixes:
* Remote read read_recent option is now false by default.
* Pass the right configuration to each Alertmanager (AM) when using
multiple AM configs.
* Fix not-matchers not selecting series with labels unset.
* tsdb: Fix occasional panic in head block.
* tsdb: Close files before deletion to fix retention issues on Windows
and NFS.
* tsdb: Cleanup and do not retry failing compactions.
* tsdb: Close WAL while shutting down.
2018-01-29 13:36:06 +01:00
2017-11-18 18:00:03 +01:00
-------------------------------------------------------------------
Thu Nov 9 07:30:09 UTC 2017 - tserong@suse.com
- Update to final v2.0.0 release (bsc#1067341):
+ Compared to version 1.x, this release includes a completely rewritten
storage engine, huge performance improvements, but also many backwards
incompatible changes. For more information, read the announcement blog
post and migration guide:
- https://prometheus.io/blog/2017/11/08/announcing-prometheus-2-0/
- https://prometheus.io/docs/prometheus/2.0/migration/
+ Changes:
* Completely rewritten storage layer, with WAL. This is not backwards
compatible with 1.x storage, and many flags have changed/disappeared.
* New staleness behavior. Series now marked stale after target scrapes
no longer return them, and soon after targets disappear from service
discovery.
* Rules files use YAML syntax now. Conversion tool added to promtool.
* Removed count_scalar, drop_common_labels functions and keep_common
modifier from PromQL.
* Rewritten exposition format parser with much higher performance. The
Protobuf exposition format is no longer supported.
* Example console templates updated for new storage and metrics names.
Examples other than node exporter and Prometheus removed.
* Admin and lifecycle APIs now disabled by default, can be reenabled via
flags
* Flags switched to using Kingpin, all flags are now --flagname rather
than -flagname.
+ Features:
* Remote read can be configured to not read data which is available
locally. This is enabled by default.
* Rules can be grouped now. Rules within a rule group are executed
sequentially.
* Added experimental GRPC apis
* Add timestamp() function to PromQL.
+ Enhancements:
* Remove remote read from the query path if no remote storage is
configured.
* Bump Consul HTTP client timeout to not match the Consul SD watch
timeout.
* Go-conntrack added to provide HTTP connection metrics.
+ Bug Fixes:
* Fix connection leak in Consul SD.
-------------------------------------------------------------------
Tue Nov 7 23:30:46 UTC 2017 - tserong@suse.com
- Update to v2.0.0-rc.3 (bsc#1067030):
+ Enhancements:
* Remove remote read from the query path if no remote storage
is configured.
* Bump Consul HTTP client timeout to not match the Consul SD
watch timeout.
* Bump up a too small max block duration to the min block
duration instead of returning an error.
+ Bug Fixes:
* Avoid needless padding of 4 zero bytes in index files.
* Delete old blocks during reload (necessary on MS Windows).
* Fix regression of alert rules state loss on config reload.
* Serialize background WAL operations to avoid data races.
+ This also contains bugfixes and remote-storage features from
the 1.8 branch.
-------------------------------------------------------------------
Fri Oct 27 02:06:17 UTC 2017 - tserong@suse.com
- Update to v2.0.0-rc.2 (bsc#1065377):
+ Enhancements:
* Handle WAL segments with corrupted header gracefully
* Stabilize memory usage during WAL replay
+ Changes:
* Prefix all storage metrics with prometheus_
+ Bug Fixes:
* Correctly handle label removal in remote read
* Fix chunk misalignment causing out-of-order samples
* Fix connection leak in Consul SD
* Handle invalid chunk dereferences gracefully
* Prevent potential deadlock during failing querier construction
-------------------------------------------------------------------
Thu Oct 19 00:00:32 UTC 2017 - tserong@suse.com
- Update to v2.0.0-rc.1:
+ Features/Enhancements:
* Added a warning for time-drift between the browser and the
prometheus-server.
* Much faster WAL read-back on restart.
+ Bug Fixes:
* Fixed Remote-read to not drop the first series.
* Validate recording-rule names.
* Fix several races.
* Only close blocks if there are no iterators accessing it.
- Refresh 0002-Default-settings.patch
-------------------------------------------------------------------
Wed Oct 11 01:13:03 UTC 2017 - tserong@suse.com
- Update to v2.0.0-rc.0:
+ Numerous changes to the new storage layer, the main changes being:
* Remove `count_scalar`, `keep_common` and `drop_common_labels` functions
* Breaking change in the index format for better consistency
* Fix panic due garbage collected mmap'ed strings
* Fix broken snapshots and admin APIs
* Send HTTP Accept header when scraping
* Use the WAL flush interval passed instead of the hardcoded value
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
- Refresh 0002-Default-settings.patch
-------------------------------------------------------------------
Sun Sep 24 15:11:37 UTC 2017 - tserong@suse.com
- Fix loopback address for REST API gateway (bsc#1059462)
+ Added pr-3174.patch
-------------------------------------------------------------------
Fri Sep 22 08:07:50 UTC 2017 - tserong@suse.com
- Update to v2.0.0-beta.5:
+ Bug Fixes:
* Remove deadlock on startup when restoring WAL
* Fix semantical races resulting in invalid persisted files
* Correctly read back WAL in certain edge cases
* Prevent crashes caused by changing metric representations in target's
/metrics
+ Enhancements:
* Overall memory usage reduction
* Serve debugging endpoints while TSDB is loading
* Healthy endpoint correctly reflects liveness during startup
* Switch to consistent usage of go-kit/log
+ This release may have issues with files written by previous beta releases
- Refresh 0002-Default-settings.patch to apply cleanly again
-------------------------------------------------------------------
Mon Sep 18 12:18:17 UTC 2017 - tserong@suse.com
- Update to v2.0.0-beta.4:
+ Numerous changes to the new storage layer, the main changes being:
* Single, compacted write ahead log
* Single in-memory block with garbage collection
* Improve compression of index files
* Cache series dropped via metric_relabel_configs
* Pool byte buffers for scraping
+ Overall the changes achieve a baseline reduction in memory consumption
and reduce peak memory usage by 30-40% compared to 2.0.0-beta.2
+ This release requires a clean storage directory and is not compatible
with files created by previous beta releases
-------------------------------------------------------------------
Tue Aug 22 08:10:18 UTC 2017 - tserong@suse.com
- Update to v2.0.0-beta.2
+ vendor: update prometheus/tsdb (Fix panic caused by 0 division)
-------------------------------------------------------------------
Wed Aug 16 02:17:42 UTC 2017 - tserong@suse.com
- Update to v2.0.0-beta.1
+ Includes a new storage layer, which reduces bottlenecks and shows
considerable performance improvements, but does not work with old v1
storage data.
- Refresh 0002-Default-settings.patch to apply cleanly again
2017-05-18 00:07:04 +02:00
-------------------------------------------------------------------
Tue May 9 05:32:44 UTC 2017 - tserong@suse.com
- Add systemd unit file and default config (/etc/prometheus/prometheus.yml)
- Run as unprivileged "prometheus" user
- Default to /etc/prometheus for config and /var/lib/prometheus for metrics
+ Added 0002-Default-settings.patch
-------------------------------------------------------------------
Wed Apr 26 09:07:42 UTC 2017 - tserong@suse.com
- Add BuildRequires: glibc-devel-static to fix aarch64 build
-------------------------------------------------------------------
Wed Mar 29 09:50:28 UTC 2017 - moio@suse.com
- Initial version
