MountVolume.SetUp failed for volume "oneagent-bin"

[shakib]

Anyone familiar with this error and how to resolve it?
MountVolume.SetUp failed for volume "oneagent-bin" : rpc error: code = Unavailable desc = version or digest is not yet set, csi-provisioner hasn't finished setup yet for tenant: 

I have dynatrace deployed in my eks using the guidance in Get started with full observability (cloud native full stack deployment) - Dynatrace Docs

[MaciejNeumann]

Hello @shakib,

In this case, it would probably be best to open a support ticket. 

[JulianRuizC]

I am deploying the dynatrace operator and following the kubernetes observability and application observability approach, when restart the application pods it's unable to start due to the error mentioned:

Unable to attach or mount volumes: unmounted volumes=[oneagent-bin], unattached volumes=[], failed to process volumes=[]: timed out waiting for the condition

MountVolume.SetUp failed for volume "oneagent-bin" : rpc error: code = Unknown desc = record not found

[gopher]

Hi @shakib ,

This is most likely because you are required to configure read-only volumes (like below).
Set this annotation value in your crd and re-apply the config and you should at least move past this error. 

Configure read-only CSI volumes - Dynatrace Docs

Thanks 

[JulianRuizC]

The issue was a bug on the dynatrace operator version 1.2.1, now using the last version 1.2.2 the issue is solved without putting the readonly csi driver flag to true on the dynakube.

[romil618]

I'm using operator version 1.2.2, I'm still getting this error,

Warning FailedMount 5s (x7 over 39s) kubelet MountVolume.SetUp failed for volume "oneagent-bin" : rpc error: code = Unavailable desc = version or digest is not yet set, csi-provisioner hasn't finished setup yet for tenant: 

I tried to set read only volume on crd to true, but that didn't help either.

[gopher]

All,

This can also occur if the code module image cannot be downloaded - this can be seen in the csi-oneagent pod log.
We were having this issue and just got to the bottom of it. 

In our case, there was an issue with the code module pull secret, which was preventing the code module image from being installed and returning this error in the application pod logs. 

Thanks  

[maurya]

hi @gopher, 

 Thanks for sharing the information around  "code module pull secret" , is this related to the secret getting created on the EKS cluster , which is base64 of the ApiToken from the SaaS portal?

[gopher]

Hi @maurya
In order to use application only - if the code-module image is taken from a private repository, then there is a hard requirement for a docker pull secret to be created and to be used for pulling the code-module image :Use a private registry — Dynatrace Docs

You will get this error if there is an issue with pulling the code-module image for any reason. 
This is typically a pull secret issue, since the connectivity to repository would have already been established for Active Gate and Operator.

Fun fact: Active Gate and Operator use a different mechanism to pull down the image from a private repository and do not require pull secrets.  Application only (Code-Modules) use a different pull down mechanism and require pull secrets - this mean that if you use code-modules then all images need to use pull secrets since it's a single config in the CRD.   

[Peter_Youssef]

Hello @shakib 

You can refer to this similar case for resolution:

• https://community.dynatrace.com/t5/Container-platforms/MountVolume-SetUp-failed-for-volume-quot-onea...

Hoping it adds value.

Regards,

Peter