Failed to invoke action method: coreui _Blobstore.readQuotaTypes, java-method: (Nexus

Hi all,
I’m new to Nexus.
I’ve setup Nexus using the Dockerfile from GitHub - sonatype/docker-nexus3: Dockerized version of Nexus Repo Manager 3 on RHEL8 using openjdk and podman. Nexus works well for me as a Proxy Registry for Openshift repositories but I’m seeing daily crashes of Nexus. In the /nexus-data/log/nexus.log log file, I found this:

2021-02-18 01:49:07,135+0000 ERROR [qtp1077044386-46]  *UNKNOWN - Failed to invoke action method: coreui_Blobstore.readQuotaTypes, java-method:
org.apache.shiro.authz.AuthorizationException: User is not permitted: nexus:blobstores:read
        at org.apache.shiro.mgt.AuthorizingSecurityManager.checkPermission(
        at org.apache.shiro.authz.aop.PermissionAnnotationHandler.assertAuthorized(
        at org.apache.shiro.authz.aop.AuthorizingAnnotationMethodInterceptor.assertAuthorized(
        at org.apache.shiro.authz.aop.AuthorizingAnnotationMethodInterceptor.invoke(
        at org.apache.shiro.guice.aop.AopAllianceMethodInterceptorAdapter.invoke(
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
2021-02-18 01:49:07,157+0000 ERROR [qtp1077044386-222]  *UNKNOWN - Failed to invoke action method:, java-method:
org.apache.shiro.authz.AuthorizationException: User is not permitted.

Am I doing something wrong? Is there a way to fix this?
Thank you,

AuthorizationException: User is not permitted: nexus:blobstores:read

No, everything seem to work as designed, except for the unneccessary log spam. The user attempted to read config of blob stores, but didn’t have permission to do so.

If you would like to see the log spam go away vote for the ticket:

That’s interesting, Thank you… so if If I’m seeing Nexus restarts it must come from some other place, then?

Correct. The logged error message is highly unlikely to be the cause.
Is it possible you’d be running out of memory?

Hi David, Thank you for your answer,
I gave the el8 VM hosting Nexus 4vcpus and 16Gb of memory. It’s only hosting 20GB of blobs at the moment. There’s about 8Gb of free RAM at any given time. What would be indicative of a restart in the logs?

Have you configured the Java heap? You can check our recommended setting in System Requirements
If you are running out of memory it would see java.lang.OutOfMemoryError logged somewhere.
Since you’re running on Kubernetes, scan your logs for Killed being logged alone on a line - that would indicate it was killed by k8s.
Rememebr you can use Prometheus to monitor your NXRM to help you find out what could be triggering your restarts.

Hi David,
I’m not running on k8s, only running Nexus as a container on a rhel8 VM.
I’m looking at the logs and cannot find an ‘OutOfMemory’ error or similar.
Likewise, the container doesn’t seem to have restarted, in fact the only time I thought there was a restart was because logging into the Nexus dashboard said something along the lines of ‘Server has been restarted’. Perhaps that’s a false positive. I’ll keep monitoring the situation but many thanks for trying to help.
Kind regards,

1 Like