[SOLVED] Upgrade NE5RC2 to NE5GA results in no Admin GUI

[kdsoukup]

After upgrading from NE 5.0 RC2 to NE 5.0 GA, I have no Web Admin GUI, just a bunch of garbled TEXT. User Web GUIs (both AJAX and HTML) work just fine.

I've been evaluating Zimbra for about 45 days now and am very keen on what I'm seeing. It's a tough crowd here, and am trying to fairly gauge all aspects of the product, including upgrades and admin expectations.

I'm running CentOS 5 x86 in expectations of running RHEL 5 (of which we're already using in-house) if we are able to proceed with a production rollout of Zimbra.

- I upgraded 4.5.9 to 4.5.10 just fine. Very nice.
- Upgraded 4.5.10 to 5.0 RC2 without a hitch (although the IM was not working at all ... but no worries mates). Again, impressive upgrades.
- I then upgraded 5.0 RC2 to 5.0 GA, no errors reported during that process. - And again, I'm still impressed Great work, All.

Only setup errors I can find are:

zmsetup...log:Moving deferred
zmsetup...log:ERROR: account.SERVER_EXISTS (server already exists: blobtest.monroeclinic.org)
zmsetup...log:ERROR: service.INVALID_REQUEST (invalid request: port 993 conflict between zimbraImapSSLBindPort and zimbraImapSSLProxyBindPort on server f.q.d.n)
zmsetup...log:ERROR: service.INVALID_REQUEST (invalid request: port 993 conflict between zimbraImapSSLBindPort and zimbraImapSSLProxyBindPort on server f.q.d.n)
zmsetup...log:ERROR: service.INVALID_REQUEST (invalid request: port 993 conflict between zimbraImapSSLBindPort and zimbraImapSSLProxyBindPort on server f.q.d.n)
zmsetup...log:ERROR: account.ACCOUNT_EXISTS (email address already exists: spam.s8l8pg1h@f.q.d.n)
zmsetup...log:ERROR: account.ACCOUNT_EXISTS (email address already exists: ham.bapd_ptor@f.q.d.n)

- I can't locate any runtime errors.
- All services appear to be running.
- Fresh 5.0 GA install on similar CentOS 5 server (VMWare Workstation) went well.
- I've stopped and restarted everything with no change.
- The ONLY change I made during this upgrade and the fresh install or previous upgrades was the addition of the Archiving option during the install process.

Any help would be greatly appreciated. We're getting close to the wire here and a successful upgrade/recovery would be helpful to the evaluation!

One last thing: we're getting ready to test aGroupwise 6.0 (mail.calendar/contacts) migration this week Any tips from the gallery?

Cheers.

[phoenix]

See if you've got the proxy server running with:

zmperditionctl status

If it is running then you should follow the instructions here to disable it and restart your server.

[kdsoukup]

Bill, thanks for the quick reply. Following your instructions, it looks like it's a nay:

$ zmperditionctl status
imap4 is not running
imap4s is not running
pop3 is not running
pop3s is not running

[phoenix]

Sorry, my mistake. For version 5 it should be this:

zmproxyctl status

[brian]

This sounds more like the skin problem others have reported.
[SOLVED] Admin UI unreadable after upgrade

Try https://server:7071/?skin=beach

We are still investigating this issue.

[kdsoukup]

Bill and Brian,

FYI:

$ zmproxyctl status
nginx is not running
memcached is not running

After reading the thread Brian pointed me to, I'm wondering how I could have missed it during my forum searches. It describes exactly the same symptoms I was seeing and the resolution (https://server:7071/?skin=beach) WORKS.

Problem SOLVED.

Is there any information I can send you regarding the install that may help with your investigation? Anything you need, I'll pass along.

Thanks a bunch for the help.

Last thing, I'm still looking for pointers that might help with a successful Groupwise 6.04 to Zimbra mail/calendar/contacts migration tomorrow (Friday). Any recommendations you can pass along would be appreciated.

Many cheers, Mates.

Kevin

[brian]

The root cause of the problem was worked out earlier today. The following will fix the problem long term.

zmmailboxdctl stop
rm -rf /opt/zimbra/jetty/work/zimbraAdmin
zmmailboxdctl start