Hi @manuharari !!
I believe this issue is related to this PR:
Note that this PR was backported only to 1.28.15:
I noticed you are using a really old version, 1.26.0
So I suggest you to upgrade path:
1.26.0 → 1.26.latest → restart → 1.27.latest → restart → 1.28.latest
in between each restart, wait all the data get loaded and the cluster stabilized.
Or, of your data is not too big, you can copy your data over to a new cluster running the latest version or the target version.
Let me know if this helps!
Thanks!