I also had this error. Rebooting resolved the issue, but if that is always the case then updates to the UPM need to include that fact in the prompt and upgrade instructions. Having it seem to be an innocuous update is misleading.

Still got this issue today (9.12.5). Over 20 minutes I waited for this app to update. What's worse though is that it seems to be spearing to other Atlassian plugins. I had this problem with updating "SSO for Atlassian Data Center" and "Jira Service Management" once the Jira Software platform had been updated.

https://getsupport.atlassian.com/browse/SDS-63169

We just upgraded to DataCenter license and are experiencing the same issue.

8.14.1 JS server and JSM 4.14.1 to DC was fine.  When we upgraded to 8.15 and 4.15, it seemed to add a bunch of system add ons and unlock alot of unnecessary languages and we can't seem to run any add on updates while not in safe mode.  We are not a clustered environment.   

Greg, I think because they'd have to list every version in Jira history

Same problem here. Why are the affects versions not listed?

Our Production had an outage early 1 hour due to updating this buggy UPM plugin. Try to fix this soon as it is was a impact.

Yea, we have the same situation that after update we need to restart all nodes not a good idea.

This just locked up our Jira instance for 45 minutes, with no recourse other than a complete restart of Jira. Not great.

What a mean bug. Nearly gave me a heart attack when monitoring gave me an alert while updating UPM on customers Jira.

Jira software 7.13.1 although affected by this. give it a fix please!

Our current procedure is to declare a planned outage of Jira for any plugin update, especially the UPM plugin. In Jira Data Center we take all the nodes out of the cluster, make the change on one node, wait for the caches to warm up again and add the nodes back into the cluster. It's about 10-15 minutes of outage that is not supposed to be necessary

This really needs to be prioritized. Please!

still, happens. in production v7.11 . How is this low prio?

I was confronted by this issue...not in dev, not in staging, but in production! It was non-peak specified maintenance hours, but still nerve-racking. I'd be a happier admin if the designated procedure was shut down Jira and update UPM offline, rather than update online and risk seeming catastrophe. 

I tried to update UPM in test instance, and also got this issue :

Referer URL: Unknown

java.lang.RuntimeException: javax.servlet.ServletException: java.lang.RuntimeException: javax.servlet.ServletException: java.lang.IllegalArgumentException: Root page 'com.atlassian.jira.gadgets:dashboard-page' does not exist!

java.lang.RuntimeException: javax.servlet.ServletException: java.lang.RuntimeException: javax.servlet.ServletException: java.lang.IllegalArgumentException: Root page 'com.atlassian.jira.gadgets:dashboard-page' does not exist! at com.atlassian.web.servlet.plugin.DynamicAuthorizationServletForwarder.forward(DynamicAuthorizationServletForwarder.java:55) [?:?] at com.atlassian.web.servlet.plugin.SanitizingServletForwarder.forward(SanitizingServletForwarder.java:32) [?:?] at com.atlassian.web.servlet.plugin.RememberingServletForwarder.forward(RememberingServletForwarder.java:51) [?:?] at com.atlassian.web.servlet.plugin.ResolvingServletForwarder.forward(ResolvingServletForwarder.java:36) [?:?] at jsp.default_jsp._jspService(default_jsp.java:68) [classes/:?]

 Atlassian, if Status is Verified, Symptom Severity is Critical, Occurrence Factor is 75%,  how can Priority be Low? Why is this issue unassigned?

 Observations in 7.2.13

• UPM update dialog box continues to spin after 20 minutes
• Refreshing the screen results in a 500 error
• Log shows Root page 'com.atlassian.jira.gadgets:dashboard-page' does not exist!

 You may be able to get back into JIRA using a Tomcat-direct port/URL and navigating to <your-jira-url-tomcat-direct-port>/plugins/servlet/upm

 PAINFUL - You will need to restart Jira to be able to use the Base URL.

I think this is the take away, but man it is scary not getting a response or log update from JIRA for so long! FYI we were on Jira software v7.7.0

Any problems are temporary. JIRA Software and bundled plugins will re-enable once UPM update is complete. This may take multiple minutes.

We ran into the same problem as well below the steps with which we managed to recover without removing plugins:

DO NOT PERFORM THESE STEPS UNLESS YOU'RE AN EXPERIENCED JIRA/SYSTEM ADMIN!!! Request Atlassian support instead!!

1. shutdown Jira
2. remove latest (newly installed) UPM version from /installed-plugins
3. clear plugin cache (https://community.atlassian.com/t5/Jira-questions/Jira-4-4-5-Jira-Startup-Failed-Plugins-did-not-start/qaq-p/443643)
4. clear tomcat cache (https://community.atlassian.com/t5/Jira-questions/Jira-4-4-5-Jira-Startup-Failed-Plugins-did-not-start/qaq-p/443643)
5. find all plugins disabled during UPM upgrade in the DB and re-enable in the DB directly (some will be system modules)
SELECT * FROM pluginstate where pluginenabled = 'false';
UPDATE pluginstate SET pluginenabled='true' WHERE pluginkey='<pluginkey from query above>';
6. start Jira

I agree with tom, if updating through the UI results in error  MOST OF THE TIMES, it would be wise to warn the users before they go clicking the "update" button.

I'm impressed the Jira QA team decided this is a "low" priority, when most users on the UPM plugin commentaries are informing the update is always buggy.

This update caused server was giving 500 and application need to be restarted with all plugins removed. This is big regression.

Same problem today.

Workaround
Do not update UPM in production JIRA applications outside of planned maintenance windows....

This warning should appear in the backend where you push the "update" - button. Not on a page you find if it's too late.....

Same issue here. Upgraded UPM from v2.20.3 to v2.22.4 on JIRA 7.4.2

+1

Hi,

This problem is not specific to the 7.3.4 version and exists for a long time. A bugfix would be much appreciated.

Regards,
Nicolas.