Master in ERROR state

1

Hi There,

I have deployed dremio on AWS EKS cluster.
All pods are up and running except master which is in ERROR state.
When did describe pod, below is the o/p:

Readiness probe failed: Get http://<loadbalancer_ip>:9047/: dial tcp <loadbalancer_ip>:9047: connect: connection refused

Could anyone please help with this?

Thanks in advance

1 Like
2

Did you get the fix? Can you pls share

3

@ajay1

If your master is not up, we need to check logs. Have you reviewed the log file via kubectl to see if there are any errors?

https://docs.dremio.com/deployment/amazon-eks/eks-admin/

4

Hi Balaji,

This is the error I got:
org.apache.hadoop.fs.UnsupportedFileSystemException: No FileSystem for scheme “dremioS3”

5

I am using type aws. My dremio.conf looks like this

{{- if eq .Values.distStorage.type “aws” }}
dist: “dremioS3:///{{ required “AWS bucketname required” .Values.distStorage.aws.bucketName }}{{ required “Path required” .Values.distStorage.aws.path }}”

6

Ignore… my above issue resolved. The core-site.xml was not matching dremio-cloud-tools/core-site.xml at master · dremio/dremio-cloud-tools · GitHub

After fixing it all pods and svc are running. However, the ELB is not responding and timing out. Which logs to check for this? To which pod does the request go to first?

7

@ajay1

Please check the master coordinator’s logfile