forked from pool/redis
Marcus Rueckert
9a6b379be3
- redis 7.2.2:
* (CVE-2023-45145) The wrong order of listen(2) and chmod(2) calls creates a
race condition that can be used by another process to bypass desired Unix
socket permissions on startup, bsc#1216376
* WAITAOF could timeout in the absence of write traffic in case a new AOF is
created and an AOF rewrite can't immediately start
* Fix crash when running rebalance command in a mixed cluster of 7.0 and 7.2
nodes
* Fix the return type of the slot number in cluster shards to integer, which
makes it consistent with past behavior
* Fix CLUSTER commands are called from modules or scripts to return TLS info
appropriately
redis-cli, fix crash on reconnect when in SUBSCRIBE mode
* Fix overflow calculation for next timer event
OBS-URL: https://build.opensuse.org/request/show/1119207
OBS-URL: https://build.opensuse.org/package/show/server:database/redis?expand=0&rev=230
README.SUSE ------------- Redis Server ============== 1. cp -a /etc/redis/default.conf.example /etc/redis/instancename.conf We use the "cp -a" here, so that our permissions are preserved. In case you copied the file with out the "-a" chown root:redis /etc/redis/instancename.conf chmod u=rw,g=r,o= /etc/redis/instancename.conf 2. change at least pidfile, logfile and dir setting # the pid file *has* to match your config filename without the ".conf" pidfile /run/redis/instancename.pid logfile /var/log/redis/instancename.log dir /var/lib/redis/instancename/ If you want to run more than one instance you also have to change the socket path and/or the ip:port combination. e.g. /run/redis/instancename.sock Also make sure if you copy configurations from somewhere, that "daemonize" should be set to "no". 3. create the database dir: $ install -d -o redis -g redis -m 0750 /var/lib/redis/instancename/ 4. systemctl start redis@instancename 5. systemctl enable redis@instancename 6. To stop/restart all instances at the same time use: systemctl restart redis.target systemctl stop redis.target Redis Sentinel ================ 1. cp -a /etc/redis/sentinel.conf.example /etc/redis/sentinel-instancename.conf We use the "cp -a" here, so that our permissions are preserved. In case you copied the file with out the "-a" chown root:redis /etc/redis/sentinel-instancename.conf chmod u=rw,g=rw,o= /etc/redis/sentinel-instancename.conf Please note that the sentinel config needs write permissions for the group. The chmod line differs from the line for the normal redis server. 2. change at least pidfile, logfile setting # the pid file *has* to match your config filename without the ".conf" pidfile /run/redis/instancename.pid logfile /var/log/redis/instancename.log If you want to run more than one instance you also have to change the socket path and/or the ip:port combination. e.g. /run/redis/instancename.sock Also make sure if you copy configurations from somewhere, that "daemonize" should be set to "no". 4. systemctl start redis-sentinel@instancename 5. systemctl enable redis-sentinel@instancename 6. To stop/restart all instances at the same time use: systemctl restart redis-sentinel.target systemctl stop redis-sentinel.target Integration with apache when using unix domain sockets ========================================================= If you plan to use redis in combination with apache, then you should add 'redis' to apache group and set 'unixsocketperm 770': $ usermod -a -G redis wwwrun $ systemctl restart apache2 then apache is able to connect to redis socket