0 Replies Latest reply on Aug 23, 2019 2:09 PM by Hobo Joe

    Wildfly Arjuna "Periodic Recovery" logs

    Hobo Joe Newbie

      I recently started receiving an excess of these logs and I'm not certain how to address them...

       

      The logs:

       

      2019-08-23 13:51:33,175 WARN  [com.arjuna.ats.arjuna] (Periodic Recovery) Transaction 0:ffff7f000101:4f3ecb1f:5d5d76c5:73c has 1 heuristic participant(s)!

      2019-08-23 13:51:33,176 WARN  [com.arjuna.ats.jta] (Periodic Recovery) XAResourceRecord restored heuristic instance: XAResourceRecord < resource:Subordinate XAResource at http-remoting://127.0.0.1:8080, txid:< formatId=131077, gtrid_length=29, bqual_length=36, tx_uid=0:ffff7f000101:4f3ecb1f:5d5d76c5:73c, node_name=1, branch_uid=0:ffff7f000101:4f3ecb1f:5d5d76c5:73f, subordinatenodename=null, eis_name=unknown eis name >, heuristic: TwoPhaseOutcome.HEURISTIC_HAZARD com.arjuna.ats.internal.jta.resources.arjunacore.XAResourceRecord@6715b993 >

      2019-08-23 13:51:33,176 WARN  [com.arjuna.ats.arjuna] (Periodic Recovery) Transaction 0:ffff7f000101:4f3ecb1f:5d5d76c5:73c restored heuristic participant XAResourceRecord < resource:Subordinate XAResource at http-remoting://127.0.0.1:8080, txid:< formatId=131077, gtrid_length=29, bqual_length=36, tx_uid=0:ffff7f000101:4f3ecb1f:5d5d76c5:73c, node_name=1, branch_uid=0:ffff7f000101:4f3ecb1f:5d5d76c5:73f, subordinatenodename=null, eis_name=unknown eis name >, heuristic: TwoPhaseOutcome.HEURISTIC_HAZARD com.arjuna.ats.internal.jta.resources.arjunacore.XAResourceRecord@6715b993 >

       

      My environment can be summarized as follows:

      - 2x virtual nodes running Wildfly 16

      - HA Cluster using domain

      - 2x non-XA datasources (postgresql)

       

      This is a lab environment, so there's no excess of traffic. The logs began to appear when I added a second ear to the servers. The new ear is a Clustered Singleton and runs a scheduled job service that queries for specific logs and if they're found uses RMI (remote interface) to the other ear in the application server (hence the reference to http-remoting on the loopback).

       

      I guess I'm curious in general about these messages because it seems like a transaction is failing that it's attempting to recover. It has appeared that some of the scheduled job results are delayed, but I could be wrong there and I haven't gathered any metrics to know that for certain. If these are nothing to be concerned over, how would I go about suppressing them?