karaf.log_c1
karaf.log_c2
karaf.log_c3
[CONTROLLER-1459] cluster : 3node auto-down time of cluster is not working properly Created: 10/Dec/15 Updated: 19/Oct/17 Resolved: 22/Aug/16 |
|
| Status: | Resolved |
| Project: | controller |
| Component/s: | clustering |
| Affects Version/s: | Lithium |
| Fix Version/s: | None |
| Type: | Bug | ||
| Reporter: | Anpukarasi Muthukumaran | Assignee: | Unassigned |
| Resolution: | Cannot Reproduce | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Operating System: All |
||
| Attachments: |
karaf.log_c1
karaf.log_c2
karaf.log_c3
|
| External issue ID: | 4751 |
| Description |
|
I observed that auto-down time function of cluster is not working properly. I created a 3 node cluster and configured 30s in my akka.conf. and I down 2 of my cluster node (shutdown the vm itself) and waited for nearly 45 mins, after that I restarted my vm (cluster) then I checked that the newly started 2 # vms (clusters) are getting formed a cluster with existing node. But as per the function, it should form a separate cluster (http://doc.akka.io/docs/akka/snapshot/scala/cluster-usage.html#Automatic_vs__Manual_Downing) Herewith I am enclosing the karaf log of all 3 nodes I found this issue in SR3 Lithium image |
| Comments |
| Comment by Anpukarasi Muthukumaran [ 10/Dec/15 ] |
|
Attachment karaf.log_c1 has been added with description: karaf_c1.log |
| Comment by Anpukarasi Muthukumaran [ 10/Dec/15 ] |
|
Attachment karaf.log_c2 has been added with description: karaf_c2.log |
| Comment by Anpukarasi Muthukumaran [ 10/Dec/15 ] |
|
Attachment karaf.log_c3 has been added with description: karaf_c3.log |
| Comment by Tom Pantelis [ 22/Aug/16 ] |
|
If I'm understanding this correctly, you stopped 2 nodes and later restarted them and they re-joined with the existing node that was still running. This is how it's supposed to work. A node always looks for another node (listed in the seed nodes) to join on startup. The documentation you refer to talks about cluster islands forming when there is a network partition between nodes. In this case, after the nodes on the other side of the partition are deemed "down", then cluster islands can/will form. |