[ovs-discuss] Infinite timeout for ovs-vsctl reconnect?

Ben Pfaff blp at nicira.com
Tue May 8 02:42:36 UTC 2012


Sounds like a good idea, thanks.

On Mon, May 07, 2012 at 07:07:11PM -0700, bkruger wrote:
> ok.  That can work..  I believe I found the source of the issue on that as
> well.   the RHEL ifup / ifdown scripts call ovs-vsctl directly w/o a time
> out set.   If you'd like I can submit a patch for this and throw in a
> default of 30 seconds or something inside it.
> 
> -b
> 
> 
> On Mon, May 7, 2012 at 2:50 PM, Ben Pfaff <blp at nicira.com> wrote:
> 
> > On Mon, May 07, 2012 at 02:49:15PM -0700, bkruger wrote:
> > > I notice that if the db is unreachable, it attempts to reconnect.   I'm
> > > looking at the code (not a C programmer, but can read some) and I see
> > that
> > > tries to reconnect infinitely.   Is there a place where this can be set
> > to
> > > give up after X attempts?   If you're on a RHEL box at least, if OVS was
> > > already stopped for whatever reason, if you try to shutdown or reboot,
> > you
> > > are stuck in an infinite loop and the only thing you can do is reset/pull
> > > the plug to reboot the machine.   I would prefer that this is some sort
> > of
> > > a settable value as opposed to editing/recompiling code for each release.
> > > Just curious if that was on the map somewhere.
> >
> > There's not currently a way to stop after X attempts, but you can use
> > --timeout to stop after a given time.
> >



More information about the discuss mailing list