Forum Discussion
I open a F5 case and a sol provided. Please refer to the following answer.
Oct 28 00:08:38 BJBJ-YZ-cBSS-HXLB02-F8950 err tmm1[13072]: 014c0001:3: diameter process ingress error Hudfilter teardown Oct 28 00:08:38 BJBJ-YZ-cBSS-HXLB02-F8950 err tmm1[13072]: 014c0001:3: diameter hud_dime_handle error Hudfilter teardown Oct 28 00:08:38 BJBJ-YZ-cBSS-HXLB02-F8950 err tmm5[13073]: 014c0001:3: diameter process ingress error Hudfilter teardown Oct 28 00:08:38 BJBJ-YZ-cBSS-HXLB02-F8950 err tmm5[13073]: 014c0001:3: diameter hud_dime_handle error Hudfilter teardown Oct 28 00:46:24 BJBJ-YZ-cBSS-HXLB02-F8950 err tmm1[13072]: 014c0001:3: diameter process ingress error Hudfilter teardown Oct 28 00:46:24 BJBJ-YZ-cBSS-HXLB02-F8950 err tmm1[13072]: 014c0001:3: diameter hud_dime_handle error Hudfilter teardown
The log entry suggests that a client side connection was terminated while the client was possibly trying to send a request. It is likely that the connection was torn down because there was a failure to load balance a message on the server side. This can be easily mitigated by following the below configuration modifications:
a custom mblb profile with "isolate-client" enabled
Create a custom mblb profile with "isolate-client" enabled. Refer to the solution article, sol13575: Configuring a BIG-IP virtual server with message-based load balancing profile
Add an iRule to ensure a message gets reload balanced in the event a selected server fails.
when LB_FAILED { if { [active_members [LB::server pool]] > 0 } { after 100 LB::reselect pool [LB::server pool] } }
If there are not too much log entry, user could just ingore it. For further analysis,you need take a capture.