To close the loop on this for anyone that digs it up later... While the issue of lock handoff is still an open one, and one that I think may require a new Locker scheme, the underlying cause of the issue I was seeing with my activemq cluster was due to a difference in understanding on how the network was supposed to be configured on the NFS server side.
I had isolated the symptom down to the fact that writes of over 16 MiB caused the NFS layer to lock up when the NFS server admins corrected the network routes on the server side, and I'm confident this has resolved the immediate issues I was seeing. Jim On Wed, Mar 23, 2016 at 7:21 AM James A. Robinson <j...@highwire.org> wrote: > Thank you for the information. Would you be able to tell me if your > NetApp settings for the lock lease are the standard values? I'm told that > it's 30 seconds, I was thinking about whether or not I needed to have that > raised much higher to prevent the particular issue I'm seeing (the tcp > layer trying to send SYN packets for 3 minutes before giving up)? > > On Tue, Mar 22, 2016 at 5:27 PM Paul Gale <paul.n.g...@gmail.com> wrote: > >> We're using the following with our NetApp device: >> >> -rw,soft,intr,noexec,nosuid,nodev,users,rsize=65535,wsize=65535,proto=tcp >> >> ​Whilst 'soft' is generally not recommended I've found that using 'hard' >> causes the broker to lock up immediately. >> The settings we have in place above were chosen by our storage team. >> >> Thanks, >> Paul >> >
