Details
-
Bug
-
Resolution: Incomplete
-
Critical
-
2.1.0
-
Security Level: Public
-
None
-
Centos
Cluster1 - 4 nodes, 6 core /guinep-s10501:8091
Cluster2 - 4 nodes, 6 core 172.23.105.58:8091
Cluster3 - 3 nodes, 4 core 172.23.121.10:8091
Cluster4 - 3 nodes, 4 core 172.23.121.7:8091
Description
Cluster1
1. Load 24M items on bucket0,
2, Setup unidirecrtional replications to cluster2, 3, 4
3. Run system over 10 hours.
4. Replicated data on Cluster3 and 4 show only 16M out of 24M items.
- Replication progress shows "100 percent complete" for all outbound replication - This is incorrect.
- Replication Queue is around 400-500k ( This is aggregated across all the replication streams) - This is incorrect, it should be at least 8+8 M items in outbound queue.
- Replication errors show " Post Condition error 500" - Can we have more detailed errors here? What is the best way to debug this error code/ message.
Remote Cluster 3, 4 stats
- Memory is nearly fully used up with Active resident ratio - 1 percent on both the clusters.
- Incoming XDCR stats show
- Zero setMeta ( which is expected)
- 5-10K get meta ops/sec
Added screenshot from Source Cluster - cluster1
and cluster3
Adding logs from the clusters.