-
Notifications
You must be signed in to change notification settings - Fork 80
CORTX-33789: Hare Config stage fail with Motr Clients #2146
Conversation
retest this please |
a786fb1
to
27a08fa
Compare
@mssawant @d-nayak @Shreya-18 please review this PR. Premerge failure is because of some infra issue. |
@pavankrishnat can you please elaborate the problem and solution in the commit message? |
38b5881
to
64950ba
Compare
@pavankrishnat request you to add appropriate commit message |
64950ba
to
18e0722
Compare
With below image:
Deployed 6N cluster without motr client successfully at https://eos-jenkins.colo.seagate.com/job/Cortx-Automation/job/RGW/job/setup-cortx-rgw-cluster/10407/console
Deployed 6N cluster with motr client successfully at https://eos-jenkins.colo.seagate.com/job/Cortx-kubernetes/job/setup-cortx-cluster-solution-input/657/console
Within data pods, able to run hare mini provisioning stages with the consul source type
It's generates hare config files successfully
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please add a comment mentioning presently, both client and data pods are returned by the conf store search query. But data pods are filtered out while creating node in
_create_node().
Tried 3N bootstrap using
|
@mssawant, Manually bootstrapped a Single Node setup with current PR branch which worked.
|
cc0a89b
to
f1d4643
Compare
Problem: If num_client_inst is specified a non-zero value, deployment is failing. Issue is with search key. For data nodes we used 'io':'motr', it returns the list of machine IDs which includes both data node and motr client nodes, because of which deployment was failing. Solution: Separated search key for data node and motr client node. Signed-off-by: pavankrishnat <pavan.k.thunuguntla@seagate.com>
a9b592a
to
addefdb
Compare
CORTX-33789: Hare Config stage fail with Motr Clients
Problem:
If num_client_inst is specified a non-zero value, deployment is failing.
Issue is with search key. For data nodes we used
'io':'motr'
, it returns thelist of machine IDs which includes both data node and motr client nodes,
because of which deployment was failing.
Solution:
Separated search key for data node and motr client node.
Signed-off-by: pavankrishnat pavan.k.thunuguntla@seagate.com