After firmware update from 11.0 to 11.1, MEP is flapping every 5 seconds. GSLB vserver status between sites do not appear to be transferred between sites.
 

The workaound is, Enable AppFlow connection chaining on the peer node before the upgrade.

Engineering is working on the permanent fix and should be available in below versions.

11.1 build 56.x and above
12.0 build 55.x and above

---

Problem Cause

Appflow connection chaining is enabled on one GSLB site and not on the other. 

set appflow param -httpUrl ENABLED -httpCookie ENABLED -httpReferer ENABLED -httpMethod ENABLED -httpHost ENABLED -httpUserAgent ENABLED -httpContentType ENABLED -httpAuthorization ENABLED -httpVia ENABLED -httpXForwardedFor ENABLED -httpLocation ENABLED -httpSetCookie ENABLED -connectionChaining ENABLED -httpDomain ENABLED -identifierName ENABLED -identifierSessionName ENABLED

- When AppFlow parameter connectionChaining is enabled on one GSLB site node, NetScaler adds 20 bytes in TCP option in the very first data packet.
- For this reason, received MSS is adjusted by 20 bytes in NetScaler
- Now when another GSLB node send MEP packet, we validated this packet size against the reduced mss and found it more than expected. So, MEP connection gets closed with reset code 9825.