Lost Pipelines after ES restart

Everytime we restart our cluster, we lose all the pipelines. So we have to push them everytime we start it again. And we should do that everyday as we power off our kubermetes cluster every night.
We have two kubernetes deployments (prod,staging) using github pires images (version 6.2.2). We are suffering this behaviour on both clusters.

Are you nodes configured to use persistent storage?

Hi,

Data nodes are using persistent storage, we don't lose the data between reboots, but master and client nodes aren't...

Master nodes also need persistent storage.

So, they are storing pipeline and cluster info?
By the way, which folders should persist between reboots?

Now I could answer my own questions...
It looks like pipeline data is stored in the master nodes, the folder that should be made persistent between reboots is /data as in the data nodes.

Thanks for your help

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.