[Gluster-devel] Regarding ETCD pod Failures on GCS setup

[Rajesh Madaka]

Hi All,

Below are the two issues has to be fixed on high priority basis.

https://github.com/gluster/gcs/issues/15 :
One of the etcd container got deleted automatically, two of the etcd
container status went to completed state #15

https://github.com/gluster/gcs/issues/29:

ETCD pods not getting spin automatically #29

ETCD pods  status went to completed state. There are so many ways to hit
this issue as mentioning below.

1) Reboot GD2 pods
2) Delete GD2 pods
3) Create PVC and mount to app pod then run I/O's on top of that mount
point. Login to GD2 pod and delete volume from GD2 pod, then try to delete
pvc from kubernetes master.
4) Create more number of PVCs like around 100 using some script.
5)  Reboot of kubernetes worker nodes

After performing above scenarios just wait for 5 to 15 mints, you will
observe that etcd pods status changing to completed.

Note: After hitting this issue we are not able to proceed with GCS setup,
GD2 pods status changing to "CrashLoopBackoff"

Regards,
Rajesh Madaka.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-devel/attachments/20181017/ec83a8a3/attachment-0001.html>