Details
-
Bug
-
Resolution: Won't Do
-
Critical
-
None
-
2.1.0, 2.1.1, 2.2.0, 2.5.1, 2.5.0
-
Security Level: Public
-
None
-
Untriaged
-
Unknown
Description
This issue was reported from one of the customers. Their cluster was extremely heavy DGM (resident ratio near zero in both active and replica vbuckets) and was highly overloaded when this memory bloating issue happened.
From the logs, we saw that the number of memcached connections was spiked from 300 to 3K during the period having the memory issue. However, we were not able to correlate the increased number of connections to the memory bloating issue yet, but plan to keep investigating this issue by running the similar workload tests.