Some additional information would clarify situation, I think. After
that failure we restarted ActiveMQ and all goes ok (we clean up space
on device). But before restart I take thread dump of broken ActiveMQ
process. I realize that there is no "ActiveMQ Data File Writer"
thread, which is present on normally working broker. If you look at
source code of NIODataFileAppender.processQueue() you will see that
this method stops execution if IOException occurred. Is there any
possibility that ActiveMQ can continue work after IOExeception occurred.
Denis Bazhenov
On Aug 29, 2009, at 11:30 AM, Denis Bazhenov wrote:
We have some strange situation. At some point IOException: "No space
left on device" occurs on server which is running ActiveMQ. We
observe following behavior: server accepts new messages but messages
not produced to consumers. Is anybody knows what's going on and how
to gracefully continue work after IOException?
Denis Bazhenov
