Cannot Access Administration Console Over Network

[pebcomputing]

Sorry in advance if this has been covered elsewhere - please point me in the right direction.

For some time now on three different zimbra servers I administer and on all versions of zimbra they've run (dating back to at least 5.0.13) the admin console sooner or later becomes inaccessible from the local network and from the WAN, however it remains accessible from the server console. I say "sooner or later" because after a zimbra restart (either via rebooting the server or an /etc/init.d/zimbra stop/start) often access works from any network location for a few minutes, a few hours, a few days... but eventually, and fairly early on, all that happens when an attempt is made to load it is a browser "cannot load this page" error due to timeout.

Time of day / server usage does not seem to be a factor. Again, zimbra version does not seem to be a factor. The three servers are running either Debian Lenny or Ubuntu 8.04, so even the server o/s doesn't seem to be a factor (although I guess it could be a "debianism"). Two of the three servers are on the same network, however the third is on a totally unrelated network, so that wouldn't seem to be a factor either.

I've been able to put up with this because I have remote desktop access across the WAN to the server console, however it's been a long time and none of the zimbra release updates has resolved it for me, so I figure either I've got a kink in my setup that I should get help straightening out, or there's an issue with zimbra that needs to be brought to the community's attention.... and it is by now getting annoying and inconvenient.

Thanks for your ideas....

[pebcomputing]

Okay, wait.... maybe it's only a problem in 6.0.x from a linux kvm instance.... I forgot I'd blocked 7071 ages ago in the firewalls since I couldn't use the admin interface remotely at that time anyway. Just realized it and unblocked it and it works except from the windows virtual machine (linux kvm based) I tend to use as for a lot of adminning.... Still the same behavior there. linux kvm networking issue, perhaps?