[CONTROLLER-1318] Clustering: Remove the member role name from the shard's persistent ID Created: 13/May/15 Updated: 18/Jan/16 Resolved: 18/Jan/16 |
|
| Status: | Resolved |
| Project: | controller |
| Component/s: | clustering |
| Affects Version/s: | Post-Helium |
| Fix Version/s: | None |
| Type: | Bug | ||
| Reporter: | Tom Pantelis | Assignee: | Kalaiselvi |
| Resolution: | Won't Do | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Operating System: All |
||
| External issue ID: | 3211 |
| Priority: | Normal |
| Description |
|
Currently we prepend the shard's persistent ID with the member role name. This presents a couple drawbacks: 1) you can’t change the member name or the shard data wont be recovered from persistence on restart. 2) you can’t backup a journal on one node and restore it on another node. We should omit the member role name and/or allow for the persistence ID to be specified in the module-shards file. To preserve data for existing installations we'll need to migrate the existing persisted shard data to the new persistent ID. Note: we need to keep the member role name as part of the shard name as we use that to identify the raft peers. |
| Comments |
| Comment by Moiz Raja [ 18/Aug/15 ] |
|
This has impact on backwards compatibility. |
| Comment by Tom Pantelis [ 18/Jan/16 ] |
|
#2 is no longer an issue with the Be online backup feature which takes an in-memory data tree snapshot and thus isn't dependent on akka persistence. So an online backup can be restored on any node. If a use case arises for changing the member name then we could add an RPC to do so. The change would have to be propagated to all nodes in the cluster (as a server config change). I'm closing this bug. |