Details
-
Bug
-
Resolution: Fixed
-
Critical
-
7.1.0
-
7.1.0-2075
-
Untriaged
-
-
1
-
Unknown
Description
- Create a 4 node cluster
- Create a 2 node XDCR remote cluster
- Create required buckets and collections.
- Create required buckets and collections on XDCR remote.
- Create 20000000 items sequentially
- Update 20000000 RandonKey keys to create 50 percent fragmentation
- Create 20000000 items sequentially
- Update 20000000 RandonKey keys to create 50 percent fragmentation
- Rebalance in with Loading of docs
- Rebalance Out with Loading of docs
- Rebalance In_Out with Loading of docs
- Swap with Loading of docs
- Stop the loading at Source side. Check XDCR items on the destination cluster. since so many hours there are continuous mutations coming but the items count are stuck at 186,879,889. Mutation reported at destination is around 50-100k/s. What are these mutations if the item counts aren't changing at the destination, Neil Huang ? Need your help in understanding.
Attachments
Issue Links
- relates to
-
MB-50240 [Magma, 2 replicas, 1%DGM]: XDCR Replication seems to be stuck on rebalance out 1 node on src cluster.
-
- Closed
-
-
MB-50074 Avg. initial XDCR rate in XDCR rebalance test dropped from 216K to 160K on build 7.1.0-1885
-
- Closed
-
-
MB-41105 UI: XDCR Backfill
-
- Open
-
-
MB-50340 XDCR - disable OSO if source bucket is using magma
-
- Closed
-
-
MB-50595 XDCR - investigate why "older node" message shows up
-
- Closed
-
For Gerrit Dashboard: MB-50416 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
169537,4 | MB-50416: prevent duplicate push merge from locking up | master | goxdcr | Status: MERGED | +2 | +1 |
169648,2 | MB-50416: ensure targetVBOpaque is populated for backfill pipeline | master | goxdcr | Status: MERGED | +2 | +1 |