Throughout this article, we will assume the namespace of your Stardog cluster is stardog-ns
, and the name of your helm release is dev-sd
. Replace this with your actual namespace and helm release.
This is an error you may get when running helm install
to install Stardog with our Helm Charts. It means that Stardog is timing out during the post-install job.
The first troubleshooting step is to check your pods, which you can do with kubectl get pods -n stardog-ns
. Your output might look like this:
> kubectl get pods -n stardog-ns NAME READY STATUS RESTARTS AGE dev-sd-stardog-0 0/1 Init:0/1 0 5m22s dev-sd-stardog-1 0/1 Pending 0 5m22s dev-sd-stardog-2 0/1 Pending 0 5m22s dev-sd-stardog-abcde 0/1 Error 0 5m22s dev-sd-stardog-fghij 1/1 Running 0 1m17s dev-sd-zookeeper-0 1/1 Running 0 5m22s dev-sd-zookeeper-1 1/1 Running 0 5m22s dev-sd-zookeeper-2 1/1 Running 0 5m22s
You can see the first Stardog pod is stuck in the status Init:0/1
. The post-install job will finish when it determines the Stardog server is up and running. In this case, it hasn't determined that, so the pod is stuck. The other two Stardog pods are Pending
because they're waiting for the first to become Ready
.
The first two troubleshooting steps to take here are to describe the first pod and to look at its logs. You can describe a pod like this:
kubectl describe pod dev-sd-stardog-0 -n stardog-ns
The output for this will be large, and the part we care about is Events
, which is at the bottom. Review this output to see why the pod is not Ready
.
To review the logs, run:
kubectl logs dev-sd-stardog-0
Ideally, the cause of error will be obvious from one of these two, but if it's not, send the output of both commands to Support.
Common sources of error
The three common sources of error here are:
- Your Stardog server can't find a license. See here on how to create a Kubernetes secret for your license.
- Your PVC can't attach to your Stardog pod. More on this below.
- Your pod can't schedule on a node because the node has insufficient resources. Try adding more nodes or more powerful nodes to allow the Stardog pod to be scheduled.
If your PVC can't attach to the Stardog pod, you can describe it like you would a pod. First, get your list of PVCs with:
kubectl get pvc -n stardog-ns
And then describe your PVC like this:
kubectl describe pvc <pvc-name> -n stardog-ns
If the cause of the error is not obvious, please send the output of this command to Support.
If you have other issues with pods not starting, please see this article.
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article