[GitHub] [activemq-nms-amqp] lukeabsent commented on pull request #59: AMQNET-656 failover stalls without reconnecting in some cases - ensure an await statement is not inside a lock statement

[GitBox <gi...@apache.org>]

lukeabsent commented on pull request #59:
URL: https://github.com/apache/activemq-nms-amqp/pull/59#issuecomment-783215340


   Had a little look, and it seems overall that there could be more Reconnect() going 
   at the same time than just one (which I believe was the original intention), and that is one risk. 
   If you suspect deadlock happening around HandleProviderError-lock(SyncRoot) then maybe 
   it is being possible cause there is provider.Close inside - though deeper inside it seems to be timeouted for 1m.
   Other risk is that HandleProviderError looks like it may be called from underlying AsyncPump (AmqpNetLite) thread (inline execution) and if it locks or waits on something, then pump stops pumping and if thing being expected is something that was supposed to be delivered by AsyncPump, then there is deadlock.
   In general, it looks like it requires some work, if I have some time I will see if I can recreate it in some test.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org