[MB-4774] compactor_died while loading data and querying view Created: 07/Feb/12 Updated: 23/Jul/12 Resolved: 09/Feb/12 |
|
| Status: | Closed |
| Project: | Couchbase Server |
| Component/s: | ns_server |
| Affects Version/s: | 2.0-developer-preview-4 |
| Fix Version/s: | 2.0-developer-preview-4 |
| Security Level: | Public |
| Type: | Bug | Priority: | Blocker |
| Reporter: | Tommie McAfee | Assignee: | Karan Kumar |
| Resolution: | Duplicate | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: | .deb build 653 | ||
| Attachments: |
|
| Description |
|
6 nodes with approx 90k docs
While loading docs and querying the view one of the servers goes offline and all subsequent requests to my view is hanging without timing out: In the diags from 10.1.2.104 I see: ** When Server state == {file,<0.26053.0>,<0.26054.0>,31672309} ** Reason for termination == ** {compactor_died, {{compact_filter_error,{error,badarith},{part_id,dups},{cbitmask,0}}, [{couch_set_view_compactor,'-compact_group/5-fun-1-',2}, {couch_btree_copy,fold_copy,3}, {couch_btree,stream_kv_node2,8}, {couch_btree,stream_kp_node,7}, {couch_btree,fold,4}, {couch_btree_copy,copy,3}, {couch_set_view_compactor,compact_view,5}, {lists,mapfoldl,3}]}} In addition, I failed over 10.1.2.104 because it went down, and when it became reachable, I attempted to rebalance it back but this task is also hanging at about 99%. Fillipe, sorry wasn't able to get debug stats for this run. |
| Comments |
| Comment by Filipe Manana [ 08/Feb/12 ] |
|
Fixed. Same fix as for |
| Comment by Karan Kumar [ 09/Feb/12 ] |
| Reopening the bug.. seeing bunch of compaction deamon dead messages. |
| Comment by Karan Kumar [ 09/Feb/12 ] |
| Latest attached files. |
| Comment by Filipe Manana [ 09/Feb/12 ] |
|
This is not same issue as before.
It is an issue in the core database Alk/Damien found recently. This error is mentioned in |
| Comment by Karan Kumar [ 09/Feb/12 ] |
|
Based on Filipe's comments Duplicate of |