Sorry yeah, IP alias. I actually think I may have found the source of my problem. I was using a connection pooler that had about max_conns connections per worker process! So despite not getting any traffic to the box the memcached connections would have been saturated.
This doesn't explain the errors in my log though... I'll keep updating with my progress thanks. D On Jun 12, 12:23 am, Vladimir Vuksan <vli...@veus.hr> wrote: > Can you send what is the command you use to invoke memcached. eth0:0 is > not really a virtual interface but an IP alias. > > Vladimir > > > > On Fri, 11 Jun 2010, David Novakovic wrote: > > We have two VPS running on linode with a virtual ethernet connection > > between them. The ifconfig info looks like the following.. (note the > > name of the interface) > > > eth0:0 Link encap:Ethernet HWaddr fe:fd:ad:e6:95:fb > > inet addr:192.168.141.205 Bcast:192.168.255.255 Mask: > > 255.255.128.0 > > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 > > Interrupt:28 > > > Memcached seems to run ok when it is listening on 127.0.0.1, as soon > > as i switch to the interface above it becomes unbearably slow! > > This means I can't run my memcached servers in a cluster. > > Even when I telnet in to do a "stats" sometimes the connection is > > closed by memcached after a long pause before it even sends a > > response. > > > When I did have each server running its own memcache instance on > > 127.0.0.1 (ie not clustered) I notice the follow coming up in the log > > right before I had to restart memcache to get it to work again. Does > > it mean anything to anyone? i notice some people have interesting ways > > to deal with this issue.. like the following node.js snippet I found > > on github:http://gist.github.com/322407 > > > r...@li159-251:~# tail -f /var/log/memcached.log > > event_add: No such file or directory > > event_add: No such file or directory > > > Thanks in advance for any insights! > > > David
