There is a problem with your JIRA access

Feraz Yulgushev_gwh6km

error when trying to update the system from 6.2.7 to 7.1.9

first, I updated to version 6.4

then to version 7.1.9

and then appeared an error "Unexpected exception during JIRA startup. This JIRA instance will not be able to recover. Please check the logs for details"

I checked database for Null. And nothing found

then, I found the inactive users and replaced them

error is not missing

I then removed the home/jira/cashes/indexes  home/jira/plugins/.bundled-plugins and .osgi-plugins

and renamed  installed-plugins

 

but all this did not help, I still see the error at startup jira. in the logs also do not see anything that could help me

Jonas Andersson

Hi,

 

I very recently did a 6.2.1 to 7.1.9 upgrade but had to step-stone it like this:

6.2.1 -> 6.3.15

6.3.15 -> 6.4.13

6.4.13 -> 7.1.9

I skipped re-indexing between the first 2 upgrades, and allowed the errors to flow, and only really took care about upgrading plugins, and in specific always starting with UPM after each upgrade, to make sure the newer plugins got installed using the right UPM. I also made sure to verify that all plugins were compatible, OR DISABLED, as i saw very similar problems to what you are describing when i didn't.

Once on 7.1.9 i removed cache dirs and rebuilt it all indexes. By not doing the reindexes between the first versions i managed to get it down to a 1.5 hour upgrade, with an added hour to reapply all customizations (server ports in server.xml, dbconfig.xml, cacerts-file).

Another gotcha i ran into was that the setenv.sh pointing to my custom keystore for some reason is ignored by JIRA 7.1.9, and instead it uses the default keystore under /opt/where.i.have.my.jira/jre/lib/security/cacerts (from the top of my head)

 

Another thing! Once the installer (runnign as root) started up the JIRA instance (using user Jira) it somehow failed badly too, with lots of errors very similar to your. Try shutting it down using the shutdown.sh script as the JIRA user, kill zombiing processes after giving it a few minutes to shut down gracefully, and then start it up again using user Jira. This allowed the server to come up. 

I also noted that giving the plugin a longer timeout could delay the startup with 30 minutes, but it actually started which allowed me to keep upgrading during my earlier attempts of upgrading by adding " -Datlassian.plugins.enable.wait=300" in setenv.sh

Hope any of this can help. Good luck!

Kind regards

Jonas