Details
-
Bug
-
Resolution: Cannot Reproduce
-
Major
-
3.0
-
None
-
Security Level: Public
-
Build 3.0.0-786
Platform = Physical
OS = CentOS 6.5
CPU = Intel Xeon E5-2680 v2 (40 vCPU)
Memory = 256 GB
Disk = RAID 10 SSD
-
Untriaged
-
Centos 64-bit
-
-
Yes
Description
Scenario:
Initial XDCR test, 5 -> 5 nodes, 1 bucket x 1G x 1KB, DGM, compaction disabled, UniDir
See attached screenshot for details.
Please notice that it decreases average throughput from ~300K ops/sec to ~200K ops/sec in this particular scenario.
The issue happens because of source nodes (downgrading destination nodes to 3.0.0-785 doesn't improve throughput).
There are gaps in memcached logs as well:
https://gist.github.com/pavel-paulau/1ab970cce224c382488b