Hi Bryan, It is indeed a defect with an asynchronous message recovery. Somehow the asynchronous recover does not recover the replicated messages.
Thanks for reporting the issue. I will have a look into what is causing it. In meantime you can try switching back to synchronous recoverer if it is feasible. I have not seen this issue with the synchronous recoverer. Kind Regards, Alex On Mon, 27 Jan 2020 at 17:25, Bryan Dixon <[email protected]> wrote: > Alex, I always use the highest durability sync policy for Master and > Replica: > SYNC. > > Also, for the virtual host, I am using this custom context variable: > use_async_message_store_recovery=true > > We use that because we used to have some queues with a large amount of > messages. I wanted to bring it up in case it matters. But we've been > using > that setting for a while and have never had this issue. > > > In node2's log file you can see the following messages after I made node2 > the master that concern me: > > 2020-01-24 11:00:00,211 DEBUG [Feeder Output for dixon01] > (c.s.j.r.i.n.Feeder) - dixon02 Feeder output thread for replica dixon01 > started at VLSN 7,024 master at 7,023 (DTVLSN:7,022) VLSN delta=-1 > socket=(dixon01(1))com.sleepycat.je.rep.utilint.net > .SimpleDataChannel@34d03b0c > 2020-01-24 11:00:00,243 INFO [StateChange-dixonbroker:dixon02] > (q.m.h.role_changed) - [Broker] [grp(/dixonbroker)] HA-1010 : Role change > reported: Node : 'dixon02' (spgmqtst2:5011) : from 'WAITING' to 'MASTER' > 2020-01-24 11:00:00,243 INFO [Queue Recoverer : app_test (vh: > dixonbroker)] > (q.m.t.recovered) - [Broker] [vh(/dixonbroker)/ms(ProvidedBDBMessageStore)] > TXN-1005 : Recovered 0 messages for queue app_test > > > See that last log message - Recovered 0 messages for queue app_test. > Isn't > that an issue? I would expect to see 50 messages recovered. Perhaps I'm > misunderstanding the log messages. > > Thanks > Bryan > > > > > -- > Sent from: > http://qpid.2158936.n2.nabble.com/Apache-Qpid-users-f2158936.html > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > >
