Hi everybody,
I'm having performance issues with the distributed cache service.
Here is my farm
WFE1&2 6Gb RAM, 4vCPU - each running web role and distributed cache role
App1 - 4GB, 4vCPU - UPS, Search, Metadata, WAS
CA - 4GB, 2vCPU - CA, search admin, smtp for sharepoint
SQL - 7GB, 4vCPU
All running WS2012 (SQL has a core installation), SQL2012
1WebApp, 40 sites collections splitted into 4 content databases, not a huge amount of data (200MB) (we've haven't gone to production yet)
The problem i have is when distributed cache (2500MB cache set) is running, every page request takes more than 6 secs.
When I turn it off (of course new social features don't work anymore when i do that), pages take less than 1 sec.
I've been trying to diagnostic this issue unsuccessfully, this is why i'm asking your help (as this service is pretty new).
I have no custom development deployed, and I already followed some best practices (BLOB caching, object caching, SQL MDOP, SQL collation, logs levels, logs on other disk...) so I think this issue is clearly related to this new service. (also we are working on this project for a while and it used to be on 2010, and very performant and fast)
Any idea?
Thanks for your help!
Cya