Galera Replication Error, Create Index on a tmp table.

[John Gentilin]

Hi,

I have a 10.0.14-MariaDB-wsrep MariaDB Server, wsrep_25.10.r4144  3 node cluster running on CentOS7
using XtraDB tables.

I believe that the health of my cluster is fine, but I am getting a bunch of the GRA*.log files in my data directory.
It seems to be related to an index create on a tmp table...

Each gra file lists this as a  error
@CREATE INDEX `tmp_ad_zone_impression_tmp_ad_zone_impression_zone_id` ON `tmp_ad_zone_impression` (`zone_id`)

I also see the corresponding error in the error log
150207  0:04:12 [ERROR] Slave SQL: Error 'Table 'revive.tmp_ad_zone_impression' doesn't exist' on query. Default database: 'revive'. Query: 'CREATE INDEX `tmp_ad_zone_impression_tmp_ad_zone_impression_zone_id` ON `tmp_ad_zone_impression` (`zone_id`)', Internal MariaDB error code: 1146
150207  0:04:12 [Warning] WSREP: RBR event 1 Query apply warning: 1, 215421789
150207  0:04:12 [Warning] WSREP: Ignoring error for TO isolated action: source: 743c068d-aa60-11e4-8ba0-364e93ccbf13 version: 3 local: 0 state: APPLYING flags: 65 conn_id: 86239087 trx_id: -1 seqnos (l: 1160107, g: 215421789, s: 215421788, d: 215421788, ts: 893893125444246)
(END)


My assumption is the tmp table did not replicate but the create index did causing the error.  I only write to a single node in the
cluster and that node is not showing the errors.. also since WSREP states it is ignoring the error and my WSRep  local state = Synced, 
I am assuming Replication is alive and well


Am I correct in my assumptions ?

Is there anything I can do about this error ?

Thank you
John Gentilin