Normally when a node is taken out of the cluster for a short period of time (for maintenance or shutdown), gcache on other nodes of the cluster help donate the missing write-set(s) when the node rejoins. This approach works if you have configured a larger gcache, or the downtime is short enough. Both approaches aren’t good, especially for a production cluster. Also, a larger gcache for the server lifetime means blocking larger disk-space when the same job can be done with relative smaller disk-space.
Re-configuring gcache, on a potential DONOR node before downtime requires a node shutdown. (Dynamic resizing of the gcache is not possible, or rather not needed now.) Restoring it back to original size needs another shutdown. So “three shutdowns” for a single downtime. No way …… not acceptable with busy production clusters and the possibility of more errors.
Given the said pain-point, we are introducing gcache.freeze_purge_at_seqno Percona XtraDB Cluster 5.7.20. This controls the purging of the gcache, thereby retaining more data to facilitate IST when the node rejoins.
All the transactions in the Galera cluster world are assigned unique global sequence number (seqno). Tracking things happens using this seqno (like wsrep_last_applied, wsrep_last_committed, wsrep_replicated, wsrep_local_cached_downto, etc…). wsrep_local_cached_downto represents the sequence number down to which the gcache has been purged. Say wsrep_local_cached_downto = N, then gcache has data from [N, wsrep_replicated] and has purged data from [1,N).
gcache.freeze_purge_at_seqno takes three values:
- -1 (default): no freeze, the purge operates as normal.
- x (should be valid seqno in gcache): freeze purge of write-sets >= x. The best way to select x is to use the wsrep_last_applied value as an indicator from the node that you plan to shut down. (wsrep_applied * 0.09. Retain this extra 10% to trick the safety gap heuristic algorithm of IST.)
- now: freeze purge of write-sets >= smallest seqno currently in gcache. Instant freeze of gcache-purge. (If tracing x (above) is difficult, simply use “now” and you are good).
Set this on an existing node of the cluster (that will continue to be part of the cluster and can act as potential DONOR). This node continues to retain the write-sets, thereby allowing the restarting node to rejoin using IST. (You can feed the said node as a preferred DONOR through wsrep_sst_donor while restarting the said rejoining node.)
Remember to set it back to -1 once the node rejoins. This avoids hogging space on the DONOR beyond the said timeline. On the next purge cycle, all the old retained write-sets are freed as well (reclaiming the space back to original).
To find out existing value of gcache.freeze_purge_at_seqno query wsrep_provider_options.
To set gcache.freeze_purge_at_seqno
set global wsrep_provider_options="gcache.freeze_purge_at_seqno = now";
Why should you use it?
- gcache grows dynamically (using existing pagestore mechanism) and shrinks once the user sets it back to -1. This means you only use disk-space when needed.
- No restart needed. The user can concentrate on maintenance node only.
- No complex math or understanding of seqno involved (simply use “now”).
- Less prone to error, as SST is one of the major error-prone areas with the cluster.
So why wait? Give it a try! It is part of Percona XtraDB Cluster 5.7.20 onwards, and helps you get IST not SST for node rejoins