krw at nobugz.com
Mon Jan 8 22:39:49 UTC 2007
Wow, thanks. Looks like it will attempt to reconnect in... 34 hours :-(
Yeah, a cap that is significantly lower than that would be a good idea.
On Jan 8, 2007, at 1:35 PM, Greg Whalin wrote:
> Assuming you have it configured to attempt a re-connect, it uses a
> falling off retry period that doubles (starting at 1 second). It
> is not capped (though it should be ... bug that it is not capped to
> something reasonable), so the retry will be pretty long after 24
> hours of outage.
> I will add a fix to cap the retry to something reasonable (I was
> actually certain that a cap used to exist at some point).
> On 1/8/07, Randy Wigginton <krw at nobugz.com> wrote:
> Don't everyone answer at once...
> My question is with respect to the Java client. What are the
> reconnect rules? If it could not connect for over 24 hours, how long
> before it tries again?
> On Jan 8, 2007, at 8:12 AM, Randy Wigginton wrote:
> > One of the uses for memcache here is to track server statistics.
> > Not mission-critical information, but "nice to have" data.
> > Over the weekend, my 1.2.0 version of memcache crashed,
> > Unfortunately the instance did not have output logging, so I don't
> > know what happened. I'll update to 1.2.1 today (should have done
> > it before anyway)
> > Question is, how long before the servers attempt to re-connect to
> > memcache? I started up the instance again about 10 minutes ago,
> > but only one machine has reconnected.
> Greg Whalin
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the memcached