Last Modified: Dec 14, 2019
Opened: Jun 14, 2019
When an HTTP or HTTP2 virtual server receives a request, it may try to send this request on an existing connection to the server. This core occurs if the clientside and serverside connections are on two different TMMs.
TMM cores, failover condition occurs, and traffic processing can be interrupted while tmm restarts.
-- A virtual server with HTTP profile is created. -- HTTP2 profile may or may not be present on the virtual server. -- An httprouter profile is present on the virtual server. -- Multiple TMMs are serving the traffic.
None. This issue is not seen with a single TMM, however, this condition may not be really enforceable.
The BIG-IP system now reuses the serverside connection when the conditions are suitable. If not a new serverside connection is created to handle the request.