Details
-
Bug
-
Resolution: Fixed
-
Major
-
None
-
None
-
None
Description
When the cluster has some changes such as, restart nodes, add/remove nodes then rebalance, throughput decreased about 50% and latency increased almost double.
Here are some example test results.
http://sdk-testresults.couchbase.com.s3.amazonaws.com/SDK-SDK/CB-4.1.2-xxxx/SvcRestartAll-HYBRID/11-16-16/066463/2008dd8c5aff930b466a1f1ae954f18a-CB.html
http://sdk-testresults.couchbase.com.s3.amazonaws.com/SDK-SDK/CB-4.1.2-xxxx/Rb2OutEpt-HYBRID/11-16-16/067340/3bd6700de6f1f615365a2b22a677f665-CB.html
To compare with other SDK,
Here is Java log
http://sdk-testresults.couchbase.com.s3.amazonaws.com/SDK-SDK/CB-4.1.2-xxxx/SvcRestartAll-HYBRID/10-27-16/061748/75cf4c070d2039b546ee5f6d6da026ff-CB.html
http://sdk-testresults.couchbase.com.s3.amazonaws.com/SDK-SDK/CB-4.1.2-xxxx/Rb2OutEpt-HYBRID/10-27-16/062193/876f386c87eafe6ae607eab709f102fe-CB.html
And LCB log
http://sdk-testresults.couchbase.com.s3.amazonaws.com/SDK-SDK/CB-4.1.2-6089/SvcRestartAll-HYBRID/11-18-16/084939/616568884062338256b217a267916466-CB.html
http://sdk-testresults.couchbase.com.s3.amazonaws.com/SDK-SDK/CB-4.1.2-6089/Rb2OutEpt-HYBRID/11-18-16/085313/e8a2cfaa7cf2106fcaa8af0b4cfa7fae-CB.html