> (2) both instance #1 and instance#2 were processing the messages, and shut down instance #1 in the middle
Are you saying that the live and backup brokers were both active at the same time? Do you have a reproducible test-case I could use to observe exactly the behavior you're describing? If not, could you create one? Justin On Thu, Aug 10, 2017 at 8:22 AM, wwang2016 <wayne.w...@impactmobile.com> wrote: > I tried to set the permission of the folder so that the user who started > the > application has full control, and the issue persisted. > > One more observation: > (1) let instance #1 create all 10000 messages, the paging folder was > created > with a paging file > (2) both instance #1 and instance#2 were processing the messages, and shut > down instance #1 in the middle > (3) instance #2 continued to process the messages until no more messages > were available to instance #2. > (4) restart instance #1 to process all messages in paging file. Instance #1 > was not able to delete the paging file > > If I shut down instance #2 before restarting instance #1, the paging file > will be accessed by instance #1 and it will be deleted after message > processing was completed. So it looks like instance #2 was accessing the > paging file, but just did not process the message. > > > > > > -- > View this message in context: http://activemq.2283324.n4. > nabble.com/paging-directory-not-accessible-to-cluster- > member-tp4729527p4729598.html > Sent from the ActiveMQ - User mailing list archive at Nabble.com. >
