[Gluster-users] AFR issue when server fails

[Keith Freedman]

I'm running: glusterfs--mainline--3.0--patch-824
which I think is pretty new.

I can possibly reproduce parts of the problem.
they're production servers so I'll have to do it during low useage 
times.  I possibly can tonight.

Keith

At 07:03 AM 1/6/2009, Krishna Srinivas wrote:
>Keith,
>Assuming that you are using one of the recent releases from 1.4, 2nd
>server's glusterfs should not have got hung and should have timed out.
>Can you easily reproduce this problem?
>Krishna
>
>On Tue, Jan 6, 2009 at 12:14 PM, Keith Freedman 
><freedman at freeformit.com> wrote:
> > ok, so I ran into another afr problem tonight.
> >
> > I have 2 servers afr-ing eachother.
> > one of them had a keniption and was in some strange half working state.
> > the other one was working fine.
> >
> > I rebooted the half working one and gluster hung on the other one.
> > it failed to timeout as expected, and just sat there until the other
> > machine was pingable. then it realized it was there but not accepting
> > connections (since it wasn't up enough to mount the filesystem
> > yet).  At this point the good server started moving along ok.
> >
> > so it seems that AFR does the right thing when the IP address is
> > connectable but the AFR process or port isn't responding, but if the
> > machine/ip is completely down, then it just hangs seemingly forever.
> >
> > ??
> >
> >
> > _______________________________________________
> > Gluster-users mailing list
> > Gluster-users at gluster.org
> > http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users
> >