pacemaker/bug-728579_pacemaker-stonith-dev-id.patch

101 lines
3.6 KiB
Diff
Raw Normal View History

commit 1e01228825eb8d4449edfbb1a1fa0c38fab4d5e6
Author: Gao,Yan <ygao@suse.com>
Date: Thu Sep 6 15:14:58 2012 +0800
Medium: stonith: Expose IDs of stonith resources to stonith agents through "$CRM_meta_st_device_id" environment variable
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
Index: pacemaker-2.0.3+20200127.cdf84f849/daemons/fenced/fenced_commands.c
===================================================================
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
--- pacemaker-2.0.3+20200127.cdf84f849.orig/daemons/fenced/fenced_commands.c
+++ pacemaker-2.0.3+20200127.cdf84f849/daemons/fenced/fenced_commands.c
@@ -901,6 +901,7 @@ build_device_from_xml(xmlNode * msg)
device->id, device->on_target_actions);
}
+ g_hash_table_insert(device->params, strdup(CRM_META "_" F_STONITH_DEVICE), strdup(device->id));
device->work = mainloop_add_trigger(G_PRIORITY_HIGH, stonith_device_dispatch, device);
/* TODO: Hook up priority */
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
Index: pacemaker-2.0.3+20200127.cdf84f849/lib/fencing/st_client.c
===================================================================
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
--- pacemaker-2.0.3+20200127.cdf84f849.orig/lib/fencing/st_client.c
+++ pacemaker-2.0.3+20200127.cdf84f849/lib/fencing/st_client.c
@@ -38,6 +38,7 @@ struct stonith_action_s {
char *action;
char *victim;
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
GHashTable *args;
+ char *dev_id;
int timeout;
int async;
void *userdata;
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
@@ -596,6 +597,7 @@ stonith__destroy_action(stonith_action_t
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
}
free(action->output);
free(action->error);
Accepting request 625813 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.0+20180726.3d81c89b8: - Shutdown corosync after a fatal error - tools: ipmiservicelogd FTBFS on ppc64le - tools: notifyServicelogEvent FTBFS on ppc64le - tools: detect CIB directory correctly - tools: avoid memory leak when crm_resource checks resource - controller: avoid any chance of use-of-NULL when doing LRM op - resources: controld doesn't need to check for corosync 1 plugin - resources: use correct variable name in Stateful - cts-scheduler: restore ability to pass arguments to single test run - resources: use ocf_is_true in SysInfo - spec: drop superfluous dependency of -libs-devel on -cts - tools: update URL in crm_simulate help - crm_mon: rhbz#1461964 - add fence history features - Update to version 2.0.0+20180706.8cf3fe749 (Pacemaker-2.0.0): - tools: Multiple -V increase debug level (#1526) - scheduler: clarify multiple-anonymous-instances message - scheduler: avoid use-after-free when on-fail=block - controller: add node name to meta-data call environment - controller: filter certain synthesized failures from CIB - controller: filter certain action timeouts from CIB - controller: don't record pending notifies in CIB - executor: cancel recurring monitors if fence device registration is lost - controller: avoid memory leak if start-up fails - XML: schemas.c: fix possible memleak + less magic with numeric constants - tools: stonith_admin -h/-H options are queries - libstonithd: avoid memory leak when getting meta-data - XML: resources-3.1.rng: relaxed schema after tight upgrade enforcement - XML: schemas.c: take severities of XSLT messages into account properly OBS-URL: https://build.opensuse.org/request/show/625813 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=307
2018-07-27 17:42:24 +02:00
+ free(action->dev_id);
free(action);
}
}
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
@@ -665,6 +667,8 @@ stonith_action_create(const char *agent,
if (device_args) {
char buffer[512];
const char *value = NULL;
+ const char *st_dev_id_key = CRM_META "_" F_STONITH_DEVICE;
+ const char *st_dev_id_value = NULL;
Accepting request 532208 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Log: Reduce verbosity of developer logging - PE: We most definitely need to probe connection resources - PE: Ensure the bundle nodes get set with the correct discovery mode - PE: Implement probing of container remote nodes - attrd: Tweak log messages for clarity - crm_resource: Do not send history erasure commands for nodes where this is none - PE: Resources in bundles should also respect failcounts - PE: Do not send notifications to unclean bundles - PE: Prevent graph loops when fencing the host out from underneath a bundle - Tools: Allow bundle resources to read/write master scores based on the same logic as attrd (cf34f4c) - PE: bundle: Send an additional host list when container-attribute-target=host - PE: Correctly interpret colocation with bundles - PE/attrd: Allow bundle resources to read/write attributes based on the physical host rather than the container name - Upstream version cs: 4328c6a65424d1caf9785107664ec1b7fc824664 - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Fix: -Wformat-truncation=2: prevent format truncation - snprintf: use correct signedness in print format specifiers - snprintf: use sizeof(X) as size + prevent off-by-ones - crmd,libcrmcommon,libcluster,tools: handle PID as string properly - tools: use right size for string buffer in fake_transition.c - crm_resource: fix a memleak - pengine: sprintf -> snprintf - libcrmcommon: fix possible infloop in buffer_print - crm_snprintf_offset like parts: fix off-by-ones - attrd: improve update messages - stonithd,lrmd,pengine: spelling/formatting corrections - build: install ifspeed resource agent - libcrmservices: avoid assert for HB resource with no parameters OBS-URL: https://build.opensuse.org/request/show/532208 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=273
2017-10-06 15:22:51 +02:00
snprintf(buffer, sizeof(buffer), "pcmk_%s_retries", _action);
value = g_hash_table_lookup(device_args, buffer);
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
@@ -672,6 +676,11 @@ stonith_action_create(const char *agent,
if (value) {
action->max_retries = atoi(value);
}
+
+ st_dev_id_value = g_hash_table_lookup(device_args, st_dev_id_key);
+ if (st_dev_id_value) {
+ action->dev_id = strdup(st_dev_id_value);
+ }
}
return action;
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
@@ -809,6 +818,10 @@ internal_stonith_action_execute(stonith_
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
svc_action->cb_data = (void *) action;
Accepting request 746348 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.2+20191106.eef56dcd6: - libcrmcommon: Make sure to free the --output-as description. - libcrmcommon: Rename formatted output cmdline options. - libcrmcommon: Group all output options together. - Update to version 2.0.2+20191031.1d695899e: - scheduler: clarify message about when watchdog will be used - scheduler: properly detect whether guest node is fenceable - scheduler: initialize unfencing digests to null - scheduler: improve graph synapse logging - Update to version 2.0.2+20191023.7a12ccbe4: - tools: Explain TIMESPEC in crm_mon help output. - tools: crm_mon --interval should understand more formats. - libcrmservice: don't close descriptors above current limit - rpm: pacemaker-cts requires pacemaker - tools: restore ability to show groups in crm_mon console mode - tools: Deprecated old-style output options in crm_mon. - cts-scheduler: test "year-2038" only runs for 64bit systems - fencer: clarify more messages - tools: don't ignore cib_new() errors in crm_mon - tools: avoid memory leak in crm_mon when displaying nodes - libpe_status: avoid memory leak when outputting resources - libcrmcommon: avoid memory leak when IPC compression fails - Security: introduce restrictive usage libxslt's security framework - libcrmservice: try not to spam close() file descriptors - tools: Support "crm_mon --as-html=<filename>". - tools: Don't output nodes if they don't have failures. - tools: Update checks around crm_mon --daemonize. - libpe_status: Use CSS to style messages. OBS-URL: https://build.opensuse.org/request/show/746348 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=331
2019-11-07 17:09:29 +01:00
set_bit(svc_action->flags, SVC_ACTION_NON_BLOCKED);
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
+ if (action->dev_id) {
+ svc_action->rsc = strdup(action->dev_id);
+ }
+
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
/* keep retries from executing out of control and free previous results */
if (is_retry) {
free(action->output);
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
Index: pacemaker-2.0.3+20200127.cdf84f849/lib/services/services_linux.c
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
===================================================================
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
--- pacemaker-2.0.3+20200127.cdf84f849.orig/lib/services/services_linux.c
+++ pacemaker-2.0.3+20200127.cdf84f849/lib/services/services_linux.c
@@ -28,6 +28,9 @@
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
#include "crm/common/mainloop.h"
#include "crm/services.h"
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
+#include "crm/stonith-ng.h"
+#include "crm/fencing/internal.h"
+
#include "services_private.h"
Accepting request 769083 from home:yan_gao:branches:network:ha-clustering:Factory - Update to version 2.0.3+20200128.df6c286d9: - scheduler: ensure attenuated scores still have some effect - scheduler: ignore colocation constraints with 0 score - Update to version 2.0.3+20200127.cdf84f849: - libpe_status: Remove extra colons from operations output. - libpe_status: correctly check for implicit resource ID collision - libpe_status: Use the same message for headers and history. - get ready for implicit -fno-common with upcoming GCC 10 (bsc#1160410) - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.3+20200116.92a6f7e97: - controller: bump feature set for shutdown-lock - scheduler: clear resource history when appropriate - tools: for down nodes, crm_resource --refresh should clear CIB only - controller: allow CRM_OP_LRM_DELETE to clear CIB only - tools: crm_resource resource checks should show shutdown locks - scheduler: display when a resource is shutdown-locked to a node - controller: don't clear shutdown locks when node rejoins - controller: mark shutdown-locked resources in resource history - scheduler: respect shutdown locks when placing inactive resources - scheduler: respect shutdown locks when placing active resources - scheduler: add shutdown lock cluster options - tools: improve error checking for crm_resource cleanup/fail commands - tools: improve crm_resource "why" messages - controller: improve messages when deleting CIB resource history - Update to version 2.0.3+20200116.ea5b06fae: - rectify thinko possibly behind spurious "process will not die" msg OBS-URL: https://build.opensuse.org/request/show/769083 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=333
2020-01-31 16:54:26 +01:00
static void close_pipe(int fildes[]);
@@ -405,6 +408,15 @@ static void
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
add_action_env_vars(const svc_action_t *op)
{
void (*env_setter)(gpointer, gpointer, gpointer) = NULL;
Accepting request 691523 from home:yan_gao:branches:network:ha-clustering:Factory - spec: no need to explicitly require the virtual dependency on libqb - Rebase: * bug-728579_pacemaker-stonith-dev-id.patch - Update to version 2.0.1+20190402.e091f4f0c: - scheduler: one group stop shouldn't make another required - libcrmcommon: compare_version: drop superfluous heap allocation - libfencing: enumerate installed RHCS-style agents alpha-sorted - xml diff'ing: prevent 'id' attribute reordering in "diff-removed" - libcrmcommon: downplay "Creating RNG parser context" as debug prio - controller: improve failed recurring action messages - controller: improve lost action logs - use common service interface for fence-agents and RAs - libcrmcommon: allow whitespace in name/value pairs - libcrmcommon: pcmk_nvpair_t should handle NULL values - Update to version 2.0.1+20190312.059e2e26b: - spec: move stonith_admin to -cli where it belongs - tools: Add crm_rule. - Update to version 2.0.1+20190311.e91ee92e5: - libfencing: Change return type on stonith_agent_exists. - scheduler: cl#5301 - respect order constraints when relevant resources are being probed (bsc#1117934, bsc#1128374) - Update to version 2.0.1+20190304.1ac166cd3: - build: Update pkgconfig files for libpe. - libp-i: Add a pkgconfig file for libp-i. - libs: Rename libpengine to libpacemaker-internal. - tools: Add new validate options to crm_resource. - Add HealthIOWait provider which checks hosts's IOWait OBS-URL: https://build.opensuse.org/request/show/691523 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=323
2019-04-04 15:06:34 +02:00
+
Accepting request 721844 from home:yan_gao:branches:network:ha-clustering:Factory:Test - Update to version 2.0.2+20190801.122c32115: - libcib: install cib_types.h - tools: Use glib for cmdline handling in crm_mon. - libcrmcommon: Handle no digits in crm_int_helper. - libcrmcommon: Allow glib cmdline parsing without formatted output. - extra: calculate #health_disk correctly in SysInfo - extra: handle run-as-user properly in ClusterMon - extra: advertise promote and demote in Stateful meta-data - extra: formally deprecate the ocf:pacemaker:pingd agent - extra: quote shell variables in agent code where appropriate - extra: use ":=" where appropriate in agent code - Update to version 2.0.2+20190725.4b4fea3eb: - Rebase: * bug-977201_pacemaker-controld-self-fencing.patch - libcrmcommon: Restore getopt behavior in stonith_admin. - controller: document the cluster-name cluster property - controller: allow configurable reaction to local node fencing - pacemakerd: make daemon exit detection messages more consistent - executor: stonith probes should fail only if previously registered - libstonithd: handle API creation errors better - spec: avoid long-broken (in a pristine tree state) "make all" - controller: panic local host if notified of own fencing - controller: clear "required" bit when disconnecting scheduler - libcrmcommon: Remove -q from global cmdline options. - tools: Allow HTML output in stonith_admin. - libstonithd: Add HTML support to fencing messages. - st_client: Reword stonith_event_text. - controld: Use XML output to check for pending status. - pengine: formatted output for html OBS-URL: https://build.opensuse.org/request/show/721844 OBS-URL: https://build.opensuse.org/package/show/network:ha-clustering:Factory/pacemaker?expand=0&rev=329
2019-08-08 18:13:22 +02:00
+ if (safe_str_eq(op->standard, PCMK_RESOURCE_CLASS_STONITH)
+ && safe_str_eq(op->agent, "fence_legacy")
+ && op->rsc != NULL) {
+ const char *st_dev_id_key = CRM_META "_" F_STONITH_DEVICE;
+
+ setenv(st_dev_id_key, op->rsc, 1);
+ }
+
if (op->agent == NULL) {
env_setter = set_alert_env; /* we deal with alert handler */