Details
-
Bug
-
Resolution: Fixed
-
Major
-
3.0.1, 4.0.0
-
Security Level: Public
-
None
-
CentOS 64 Bit
Build 968
-
Untriaged
-
Centos 64-bit
-
No
-
Mar 9 - Mar 27
Description
1. Create a 4 node cluster 20,21,22,23
2. Rebalance the cluster
3. Add 200000 documents equally across cluster
4. Failover 2 nodes (20,23)
5. Index query on master node - 21
6. Rebalance with following command:
password=password&ejectedNodes=ns_172.23.107.23ns_1%40172.23.107.20&user=Administrator&knownNodes=ns_1%40172.23.107.21%2Cns_1%40172.23.107.22%2Cns_1%40172.23.107.23%2Cns_1%40172.23.107.20
8. Create ddoc - node 21
9. Remove on the ddoc created in step 8
10. Check changes in the queue - ep_queue_size
Timeout occurs during checking of the queue_size
Attachments
Issue Links
- is duplicated by
-
MB-19440 [windows] disk write queue does not go down to zero when no load running on cluster
- Closed
For Gerrit Dashboard: MB-13155 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
47580,7 | MB-13155: Decrement dirtyQueue size after rollback | sherlock | ep-engine | Status: MERGED | +2 | +1 |
47655,1 | MB-13155: Decrement dirtyQueue size after rollback | master | ep-engine | Status: ABANDONED | 0 | 0 |