On Tue, Mar 1, 2016 at 7:02 AM, artnaseef <a...@artnaseef.com> wrote:
> So something is very wrong then. NFS should *not* allow two NFS clients to > obtain the same lock. > > Three possible explanations come to mind: > > * The lock file is getting incorrectly removed (I've never seen ActiveMQ > cause this) > * There is a flaw in the NFS locking implementation itself > * The NFSv4 timeout for the lock is overly aggressive (perhaps less than 1 > second even? The timeout needs to be at least a couple of minutes in most > scenarios to be reliable) > Hrm, I'm pretty sure the server timeout is currently set to something lower, like 15 seconds. I'm not sure what the timeo is on the client side by default. Do you recommend both sides have minute-long values? Or were you discussing one particular side (server or client)? If the servers are Linux servers, try using strace to see their interactions > with the lock file. > I was going to try running a packet capture to record NFS level activity, assuming wireshark had a decoder for NFS v4 traffic. > BTW, each broker is running on its own dedicated server, right? > Yes
