Get logs from virtual server traffic

Hi Alian, There are various methods for collecting connection related data. Every five minutes sounds like a TCP timeout issue on the surface. If it is TCP, this particular method would offer some useful diagnostic data:

http://support.f5.com/kb/en-us/solutions/public/13000/200/sol13223.html?sr=37748182

Kevin

You are absolutely right. It is a TCP reset issue. Here is the error: May 29 10:08:31 VTI-PIE5-LB03 err tmm[9052]: 01230140:3: RST sent from 172.29.243.26:3080 to 172.29.240.231:2040, [0x16ba965:945] {peer} Flow expired (sweeper) (idle timeout) May 29 10:08:31 VTI-PIE5-LB03 err tmm[9052]: 01230140:3: RST sent from 172.29.243.26:2040 to 172.26.68.21:3080, [0x16ba965:945] Flow expired (sweeper) (idle timeout) I can see that the BIG-IP is resetting its connection to both server node and client. Is this normal? How can I eliviate that?

Hi Alain, The TCP profile's "Idle Timeout" could be lengthened which would consume more memory as it holds connections open. Be careful about disabling this, but it is possible as a last resort. Tweak this a bit to alleviate some of these resets first? It would seem the best option however would be to have client or server enable keep-alives.

Kevin

is keep-alive setting in tcp profile helpful?

sol8049: Implementing TCP Keep-Alives for server-client communication using TCP profiles

It's worth a try and seems like a harmless option. I was thinking that setting stops remote device from sending reset to BIG-IP. In this case, seems BIG-IP is resetting due to not receiving traffic from remote(s)? Each side should probably have keep-alive enabled.

Kevin

Thank you for your advices. I will try keep-alives to see if it solves the issue

Regards, Alain