Kubernetes cluster throwing too many "BufferOverflowError" messages.

Problem:

Kubernetes cluster throwing too many error logsBufferOverflowError, error=\"buffer space has too many data\" and buffer overflow action=:throw_exception"

Resolution:

The root cause of the issue is that there are too many nodes running inside the K8S cluster which is causing this issue.  To resolve this issue you need to increase your "replica count" and "number of threads" in your values.yaml file.

STEP: 1

1).

Increase the number of replicasCount from 3 to 6. In default values.YAML file you will find this on-line number 165.

FROM:

replicaCount: 3

TO:

replicaCount: 6

2). 

Increase the number of threads from default 8  to 16. In default values.YAML you will find this on-line number: 116

FROM:

numThreads: 8

TO:

numThreads: 16

STEP:2

Once you made this change run the below command to upgrade the helm with the new values.YAML file value.

2. Upgrade the helm using this link

$ helm upgrade collection sumologic/sumologic --reuse-values -f <downloaded-values.yaml-file>