Scans are failing,Elascticsearch cache corruption

We are analyzing a SonarQube to be used for a code scans in our company.
Scans are working fine, but suddenly from time to time all scans start failing. We are able to resolve an issue by deleting Elasticsearch cache and restarting the application.
However, multiple restarts are not possible in the real production environment.
Please, suggest, how to prevent Elasticsearch corruption in our env and stop random failures completely?
Community Edition
Version 8.9.6 (build 50800)


Welcome to the community!

What do your server logs say? Anything interesting? Any errors?


I found something interesting in the web.log. Around a failure time, there are 3393 logins for the same user:
grep web.log|grep " 08:"|wc -l
Example entry:
2022.01.31 08:59:31 DEBUG web[AX6g+6aOt0wkIvzKARtq][auth.event] login success [method|BASIC_TOKEN][provider|LOCAL|local][IP||][login|]
This user uses SonarLint plugin for IntelliJ. Plugin version 5.0.1.
Do you have any info regarding SonarLint plugin issues?


Can you share the time frame in which you saw these 3393 (literally? :exploding_head:) logins?


Hello @Bharathnidra,

It might not cause a big difference but the user is using a pretty outdated version of the plugin. I encourage them to update to the latest version (6.4.3 at the time of writing).

Also something to note is that SonarLint polls smart notifications from the server each minute. This can cause the amount of logins you observe

Hi, it happens from 8 till 9 AM, probably when user starts his work day.

1 Like

You are right, 60 logins per hour, that’s the number I see in the log per most users… Still a lot, considering thousands developers in a big organizations.


So are you saying that you observe 60 logins per user per hour, except for one user for which you observe 3000+ logins in a single hour ?