Added flow ExpandCluserWithDetectedPeers #831
Conversation
This flow would be invoked from tendrl-node-agent on the additionally detected new peers of the cluster. New nodes would be provisioned with monitoring etc and imported into the cluster in tendrl system. tendrl-bug-id: Tendrl#805 Signed-off-by: Shubhendu <shtripat@redhat.com>
|
@r0h4n @nthomas-redhat @GowthamShanmugam @a2batic @gnehapk plz review |
| } | ||
| _cluster.save() | ||
|
|
||
| node_ids = self.parameters.get("Node[]", []) |
There was a problem hiding this comment.
We should be able to detect which nodes are new (check :https://github.com/Tendrl/commons/blob/master/tendrl/commons/flows/import_cluster/__init__.py#L40)
Also, Add a flag "is_managed" to cluster node context, set this from tendrl-gluster-integration service, use this to figure out which nodes are new ("clusternodecontext.is_managed") wouldnt exist for such nodes.
There was a problem hiding this comment.
I thought detection code would be separate and after detection, it would invoke this flow. Anyway will try to introduce this is_managed flag with cluster_node_context object and use it.
There was a problem hiding this comment.
I still feel flow should be independent of detection code and as expected it should expect a nodes list for executing the expand of cluster. If that was the case we could just formed and saved jobs while detection itself, why have a separate flow defined for this?
There was a problem hiding this comment.
I am not talking about detecting the sds or making detected_cluster_id. This flow will be called only after the new nodes have the same TendrlContext (i.e same integration_id). So all you need to do is read list of nodes from here https://github.com/Tendrl/commons/blob/master/tendrl/commons/flows/import_cluster/__init__.py#L38
and then import those nodes for which ClusterNodeContext.is_managed is "no"
| enabled: true | ||
| inputs: | ||
| mandatory: | ||
| - "Node[]" |
Codecov Report
@@ Coverage Diff @@
## master #831 +/- ##
=========================================
Coverage ? 73.32%
=========================================
Files ? 85
Lines ? 3254
Branches ? 408
=========================================
Hits ? 2386
Misses ? 811
Partials ? 57
Continue to review full report at Codecov.
|
|
|
||
| loop_count = 0 | ||
| # Wait for (no of nodes) * 6 minutes for import to complete | ||
| wait_count = len(node_ids) * 36 |
There was a problem hiding this comment.
@r0h4n node_ids is not the list of nodes with which the cluster is expanded. Rather this is the full list of nodes as few skips might happen at L#70. We should wait for only no of nodes which actually executed import cluster flow here.
This flow would be invoked from tendrl-node-agent on the additionally
detected new peers of the cluster. New nodes would be provisioned with
monitoring etc and imported into the cluster in tendrl system.
tendrl-bug-id: #805
Signed-off-by: Shubhendu shtripat@redhat.com