Hi, I'm trying to set up MySQL Replication enviroment . glance around mysql RA, it uses "crm_master" to deter the promotion of the out-of-sync slave node.
https://github.com/ClusterLabs/resource-agents/blob/master/heartbeat/mysql#L496 elif ocf_is_ms; then # Even if we're not set to evict lagging slaves, we can # still use the seconds behind master value to set our # master preference. local master_pref master_pref=$((${OCF_RESKEY_max_slave_lag}-${secs_behind})) if [ $master_pref -lt 0 ]; then # Sanitize a below-zero preference to just zero master_pref=0 fi $CRM_MASTER -v $master_pref fi mysql RA sets "0" as the attribute score using crm_master. but, if there is only one node (this node might be out-of-sync), it seems that Pacemaker ignore its socre and this node will be promoted even as it is out-of-sync. I think mysql RA should be set "-INFINITY" in this case. Is it correct? I found the following comment in drbd RA, but this issue regarding negative master scores has already been resolved in the latest Pacemaker. http://git.drbd.org/?p=drbd-8.3.git;a=blob;f=scripts/drbd.ocf;h=06ffed1dc5cbf6db4c2809d33b4f40da2215d628;hb=HEAD#l410 410 *) 411 # Our local disk is not up to date. 412 # Our peer is not available. 413 # We have no access to useful data. 414 # DRBD would refuse to be promoted, anyways. 415 # 416 # set_master_score -INFINITY 417 # Too bad, at least with current (Oktober 2009) Pacemaker, 418 # negative master scores cause instance stop; restart cycle :( 419 # Hope that this will suffice. 420 remove_master_score Regards, Junko IKEDA NTT DATA INTELLILINK CORPORATION _______________________________________________ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
