Sonarqube container started to go down randomly

hello,
hoping that all is well. i am very new with sonarqube, so please excuse if my question would sound very basic at times.

we are using version 9.3 developers edition and running both sonarqube and ms sql using containers. we are using an ubuntu linux instance with docker. the VM in azure is a standard F4s v2 with 2 CPUs and 4Gig RAM. couple of plugins, AAD and Groovy are installed. we have around 54 users in the security list so far.

its been running well for the last 2 months but 2 days ago, we noticed that the sonarqube container would randomly go down. it happened at least 8 times until yesterday.

bashed into the sonarqube container and saw couple of errors from the ce.log file (see below).

2022.03.02 16:05:59 ERROR ce[][o.s.c.t.CeWorkerImpl] Failed to pop the queue of analysis reports
org.apache.ibatis.exceptions.PersistenceException:
### Error querying database. Cause: com.microsoft.sqlserver.jdbc.SQLServerException: A timeout occurred while waiting for memory resources to execute the query in resource pool 'internal' (1). Rerun the query.
### The error may exist in org.sonar.db.ce.CeQueueMapper
...
### Cause: com.microsoft.sqlserver.jdbc.SQLServerException: A timeout occurred while waiting for memory resources to execute the query in resource pool 'internal' (1). Rerun the query.

so far, we just configured the docker update to auto restart the failing container. this “safety net” seems to be working every time the sonarqube container fails BUT if the MS SQL fails, even if it auto restart, sonarqube is complaining that login to the database is failing.

our real issue here is find out why did sonarqube starts failing randomly? is it because of the error that we list above?

would really appreciate your expertise here. thanks much.

Hi,

Welcome to the community!

I would certainly start with this.

Do you have SonarQube and SQL Server on the same underlying box? Are you sure both processes have enough memory? Google is showing me that this error message ultimately comes from SQL Server, so it seems that the DB is under pressure for memory. The SonarQube failures are only the symptom; the problem is at the DB. Can you do some monitoring of what’s available to the DB when these problems happen?

 
Ann

hi ann,
thank you for a warm welcome :slight_smile:
yes, both SQ and SQL are hosted on the same VM instance of ubuntu in azure.
when i installed SQ i saw an article saying that the minimum memory requirement is 2Gig, i thought our setup should be ok because the VM has 4Gig of RAM.
alright then… i will check the available memory for SQL and would go from there.
thanks much.

1 Like