Bitbucket Server - "A fatal error has occurred"

Stefan Prelle July 14, 2017

I was successfully running BitBucket 4.14.2 and tried updating to 5.2.0 using the installer. Afterwards it complained about the old Git version, so I upgraded git to 2.11. When starting up now I get a lot of errors like

"Application context initialization for 'com.atlassian.integration.jira.integration-plugin' has timed out waiting for (|(objectClass=com.atlassian.applinks.api.ApplicationLinkService)(objectClass=com.atlassian.applinks.api.EntityLinkService))"

 

I tried to switch back to 4.14.2, but have the same effect.

I wiped the installation directory of 5.2.0 and installed new (adapting the set-*.sh scripts), but I still get those time-out errors.

When opening in a browser all I get is:

A fatal error has occurred

The following problems occurred, which prevents Atlassian Bitbucket from starting:

  • Required plugin com.atlassian.crowd.embedded.admin has failed to start
  • Required plugin com.atlassian.plugins.atlassian-nav-links-plugin has failed to start
  • Required plugin com.atlassian.bitbucket.server.bitbucket-rest has failed to start
  • Required plugin com.atlassian.bitbucket.server.bitbucket-git has failed to start
  • Required plugin com.atlassian.applinks.applinks-plugin has failed to start
  • Required plugin com.atlassian.upm.atlassian-universal-plugin-manager-plugin has failed to start
  • Required plugin com.atlassian.stash.ssh-plugin has failed to start
  • Required plugin com.atlassian.activeobjects.activeobjects-plugin has failed to start
  • Required plugin com.atlassian.support.stp has failed to start

Any Idea what I can do?

 

BTW: I tried Atlassian support, but since I am using a starter license, they deflected me to the community.

 

Update:
I deleted the database, the $BITBUCKET_HOME and $BITBUCKET_INSTALL directories (of course I made backups) and rexecuted the installer and I still get the same timeout errors and this fatal error. So I assume the problem is not a broken installation, but something alongside Bitbucket, the server itself or the network. But I don't know how to dig deeper.

 

Update 2: (2017-07-17)

What seems to help is to increase the plugin in $BITBUCKET__INSTALL/bin/_start-webapp.sh.
Following hints from similiar questions I added

JVM_SUPPORT_RECOMMENDED_ARGS="-Datlassian.plugins.enable.wait=300"

in the script and restarted. This time I was able to begin the installation wizard.

The remaining question is "Why is this powerful hardware (12 Cores, 32 GB RAM) in need of this high timeout?"

1 answer

0 votes
Christian Glockner
Atlassian Team
Atlassian Team members are employees working across the company in a wide variety of roles.
July 14, 2017

Hi Stefan,

It could be that the plugin cache got corrupted. The easiest thing to try is to clean that cache using the instructions for Entry 1 in the Resolution section of Bitbucket Server failed to Start due to Required plugin failed to start and to see if this helps you start Bitbucket Server again.

Cheers,

Christian

Premier Support Engineer

Atlassian

Stefan Prelle July 14, 2017

Hi Christian,

thanks for the reply. I deleted the directories as indicated in your link and restarted again. Unfortunately that did not help.

BTW: Startup time until the timeout error show in the log is nearly 20 minutes.

Regards,
  Stefan

Stefan Prelle July 14, 2017

Update:
I deleted the database, the $BITBUCKET_HOME and $BITBUCKET_INSTALL directories (of course I made backups) and rexecuted the installer and I still get the same timeout errors and this fatal error. So I assume the problem is not a broken installation, but something alongside Bitbucket, the server itself or the network. But I don't know how to dig deeper.

Regards,

Stefan

Christian Glockner
Atlassian Team
Atlassian Team members are employees working across the company in a wide variety of roles.
July 14, 2017

Hi Stefan,

if after starting from scratch the error persists your machine may be underpowered - what are the specs of the machine?

Also, are you running this on AWS or Google Cloud by any chance?

Cheers,

Christian

Premier Support Engineer

Atlassian

Stefan Prelle July 14, 2017

Quad Core 2.5 GHz, 32 GB RAM

CentOS in VMWare

Christian Glockner
Atlassian Team
Atlassian Team members are employees working across the company in a wide variety of roles.
July 14, 2017

Hi Stefan,

That definitely seems beefy enough - do you have any <BitbucketInstallationDir>/log/catalina.out file that you can share?

Cheers,

Christian

Premier Support Engineer

Atlassian

Stefan Prelle July 14, 2017

Hi Christian,

I do have a catalina.out (https://de.pastebin.ca/3842948) and a

/var/atlassian/application-data/bitbucket/log/atlassian-bitbucket.log

which I uploaded to

https://filebin.ca/3TO1hh6qB1vk/atlassian-bitbucket.log

 

I will investigate if it is possible that the VMWare host is overbooked, leading to performance degradation, but this has to wait until next week.

Regards,

  Stefan

Stefan Prelle July 17, 2017

The problem seems not hardware related. I upgraded to 12 virtual cores, but still got these problems.
And the VMWare host isn't configured to allow overbooking, so we are talking about dedicated CPUs.

datasecmx October 25, 2017

Well , I got same issue migrating from 5.4 to 5.5.

Tried your solutions and no fix yet.

Suggest an answer

Log in or Sign up to answer
TAGS
AUG Leaders

Atlassian Community Events