Test jira gives me "Oops, you've found a dead link."

[Chris Schulte]

I am updating the jira part of our jira service desk from 8.6.1 to 8.8.1. I'm using a postgres 11.7 database on centos 7.

I'm trying to create a duplicate test environment for jira following these instructions:
https://confluence.atlassian.com/adminjiraserver/creating-a-test-environment-for-jira-966063324.html

I created the replica database from our public instance. I created a user named jira on the test platform (same as production) and then used that user/role to create the jiradb database (also the same as production). I originally used the postgres user to create the database, with the same result. I loaded the replica database onto the test server with "sudo -Hiu jira psql jiradb < jiradb.sql."

I then reindexed and made sure that there was nothing in clusternode.

I start jira and there are no errors on the console but when I access it, I get "Oops, you've found a dead link." Is there something I need to change in the database file, like globally replace production.server with test.server? Is there a way to change the base url in jiradb.sql?

I've also noticed that when I start jira on the test server, there is only one java process running whereas there are three java processes running on my production server. Does that indicate anything?

These are the errors I see in atlassian-jira.log from startup through usage without the :

2020-04-29 07:16:14,517-0500 http-nio-8080-exec-25 ERROR anonymous 436x2x1 - 10.130.210.221,10.130.192.35 /jira [c.atlassian.instrumentation.Instrument] Unable to snapshot thread local operations (implementation of OpTimerFactory is not a ThreadLocalOpTimerFactory): null
2020-04-29 07:17:02,882-0500 ThreadPoolAsyncTaskExecutor::Thread 9 ERROR [c.a.p.s.scanner.extension.ServiceExporterBeanPostProcessor] Unable to register bean 'xmlService' as an OSGi exported service
2020-04-29 07:18:32,412-0500 Caesium-1-3 ERROR [c.a.jira.startup.IndexRecoveryLauncher] Issue index is out of date with the database by more than 24 hours. Automatic recovery is not attempted

Thanks,

Chris

[Moses Thomas · Accepted solution]

@Chris Schulte   i suggest you clear plugin cache  for a start

firstly , Please perform the following:

Shut down JIRA SD
Kill process ( ps aux | grep jira then kill -9 pid)
Delete the following hidden plugin cache directories:

JIRA_HOME/plugins/.bundled-plugins
JIRA_HOME/plugins/.osgi-plugins

Restart JIRA SD (These directories will be recreated on JIRA reboot with new plugin cache)

secondly, ensure that you have enough RAM on the Machine

[Chris Schulte]

Thanks @Moses Thomas.  

I tried that since it was suggested in another community post. I even rebooted the server. It's still giving me the oops message.

[Moses Thomas]

@Chris Schulte  Do you have enough RAM on this  server ?

[Chris Schulte]

Maybe not? This is a general development server on a VM. I've never checked its RAM before.

Development:

total used free shared buff/cache available
Mem: 1.8G 292M 1.2G 26M 324M 1.3G
Swap: 2.0G 175M 1.8G

Our production is using at least twice that

Mem: 3.7G 3.4G 104M 39M 179M 57M
Swap: 3.9G 2.2G 1.7G

[Moses Thomas]

@Chris Schulte  Increase memory  to be same as production and try again.

[Chris Schulte]

I'm contacting the person who manages the VMs now. I'll let you know how it goes. Thanks!

[Chris Schulte]

@Moses Thomas  I increased the memory but I still get the same thing.

Thank you very much for trying.

[Moses Thomas]

@Chris SchulteI Think you would need to fix indexing completely

Manual index recovery

Troubleshoot Lucene index corruption in Jira server 

[Chris Schulte]

@Moses Thomas It turned out that the issues I was having have something to do with the reverse proxy setup on this particular server. It might have something to do with the fact that, this being a general test bed, there are a number of applications running in /var/www/html.

I opened up port 8080 on the firewall and now I can run jira directly by IP address. 

I want to thank you for your effort. You showed me a lot about jira.

[Moses Thomas]

@Chris Schulte  I am very happy that you are able to resolve this Issue, please  mark  the issue as resolved as ''Firewall" was the cause  in this  case, I am sure this will be help-ful to other's  with  similar issue.