JIRA userinterface problems (also with confluence)

[Norbert]

We decided using Jira but since the decission it start to act weird (is this intended??)

the problem is, that the connection often timed out, or i don't geht any usefull connection
Often i get the message "connection reset by server...", or the website don't load completly or is garbled..

I'm quite sure the problem is not only jira related because confluence running on the same server but in its own Tomcat instance is behaving the same way.
Also this problem is Browser independet (tested with Opera & firefox 6.0)

JIRA installation is on a VMware-ESX virtual Image (Windows Server 2008R2 with 4Gb RAM, using PostgreSQL as DB)

What could i do to narrow down the problem, network and Server seem to run stable

[Norbert · Accepted solution]

Thanks for the help

The problem was a missconfiguration of the IP Adress.
Somehow the IP-Address was double assigned. After fixing it, everything is ok

[Radu Dumitriu]

AFAIK, there were some problems with network timeouts on VMware-ESX 4.0, so I'm afraid you will need to upgrade your virtual machines. This is one of the reasons I prefere dedicated servers, btw, and use the virtualization stuff for not-so-long-lived servers ....

A disclaimer though: I'm just a programmer, not a sysadmin

[Nic Brough -Adaptavist-]

Are you running it behind another web server? Apache or IIS for example?

It's *very* common to run it behind Apache, and the things you are seeing are symptoms I've seen when the Apache and Tomcats are not talking to each other properly, so I thought it's worth asking to rule it out.

[Norbert]

no i installed it as like in the "Here Be Dragons" guide

Stage 1 to 3 (we don't use Bamboo, Fisheye and the other stuff)

[Norbert]

There is no firewall betweent the servers....
hmm, the question is, why JIRA have a problem with this, the other service work fine (AFAIK)?

[Norbert]

There is know firewall between active Directory(LDAP) and the Jira server

[Radu Dumitriu]

Of course. This may come for a number of reasons (firewalls, etc). You correctly configure firewalls, see for instance:

https://jira.atlassian.com/browse/CWD-568

[Norbert]

Hi I just hat an error an pasted it to the official Jira Support

I have erros like
2011-08-23 11:07:05,383 http-8080-5 ERROR [500ErrorPage.jsp] Exception caught in 500 page null
com.atlassian.crowd.exception.runtime.OperationFailedExceptio
or more intresting

2011-08-23 08:22:20,696 http-8080-2 ERROR anonymous 502x8030x1 9vm6cx 10.1.1.176 /rest/gadget/1.0/login [crowd.manager.application.ApplicationServiceGeneric] Directory 'Delegated LDAP Authentication' is not functional during authentication of 'xxxxxx'. Skipped.
2011-08-23 08:22:20,696 http-8080-2 ERROR anonymous 502x8030x1 9vm6cx 10.1.1.176 /rest/gadget/1.0/login [jira.security.login.JiraSeraphAuthenticator] Error occurred while trying to authenticate user 'xxxxxx'.
com.atlassian.crowd.exception.runtime.OperationFailedException
    at com.atlassian.crowd.embedded.core.CrowdServiceImpl.convertOperationFailedException(CrowdServiceImpl.java:869)
...
Caused by: org.springframework.ldap.ServiceUnavailableException: insv00915:389; socket closed; nested exception is javax.naming.ServiceUnavailableException: XXXXX:389; socket closed; remaining name 'OU=v2c2,DC=XXXXX,DC=XX'
    at org.springframework.ldap.support.LdapUtils.convertLdapException(LdapUtils.java:210)

It looks like the LDAP server has some problems..
could this cause such problems?

[Radu Dumitriu]

Norbert, without the logs (if this is Jira related) noone can help you. So please scan the Atlassian logs for errors and post them here

[Norbert]

This is ok because the Server are in another building on the other side of town.

[Radu Dumitriu]

11ms for a local network is way too much. Check switches, connectivity, etc.

[Norbert]

Hmmm...
i let the ping run vor some time

Statistic:

Pakete: send= 5722, received= 5702, lost = 20
(0% Verlust),
time in millisek:
Min= 2ms, Max= 68ms, Average= 11ms

could this be the problem?

[Norbert]

Usage on server (the Image or the real Hardware)
on the Image, we only have 7 Users where most of the people are currently on holiday or work on some other stuff.
So there are not more than 2 people working in parallel (at most), IMHO is the serverload "non existing"

The DB runs on the same System, so also no load on the system.

[Radu Dumitriu]

Ok, Jira logs ? Tomcat outputs ? Usage on the server ? Could you run in some sort of database pool contention ? (like for instance you have 10 connections to the DB and 1000 users? ) Thread contention ?

[Norbert]

I tried pinging the server for 1 hour, not a single package was lost.
I know its not a 100 % network test but IMHO is the problem not in the network connection..

[Norbert]

The Image is running on VMWare Server ESX 4.1