Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. If you'd prefer to explore, try our test area to get started. And see here for current known issues.

New Member

Questions on ACE operation

We are deploying a Data Center with Cat 6500, FWSM and ACE (in bridged mode, for layer 4 load-balancing between http servers), and we are experiencing intermittent performance issues (the system is not yet under production).

During our investigation, we captured traffic before and after the ACE and noticed the following things :

¤ Half of sessions are initiated from the standby ACE to the real servers whereas FT config and status seem ok. Is this normal ?

¤ An analysis software reports excellent response times (0.15 ms) but too fast TCP retransmissions (between the ACE and the real servers); displaying the connection dynamics, we see that each TCP packet (SYN, ACK, etc.) is sent at least 3 times consecutively. Is this normal behavior ?

Besides, we configured TCP probes with a 20s interval. When a Web server comes up, it takes the ACE too much more than 20s to consider it up. Is this normal ?

Note : http probes showed to be somewhat unstable.

Thank you for your help.


Re: Questions on ACE operation

the default or sdtandard is is 60 sec , if you like you can select the fast one which is having 4 sec interval.

New Member

Re: Questions on ACE operation

What version of code are you running, I would look for bugs, etc. early versions of the code were very buggy and almost unusable.

New Member

Re: Questions on ACE operation


He have just upgraded due to the instability of the SMTP probes. Here is an extract from sh version :


loader: Version 12.2[120]

system: Version A2(1.1) [build 3.0(0)A2(1.1) adbuild_00:25:02-2008/06/05_/a


system image file: [LCP] disk0:c6ace-t1k9-mz.A2_1_1.bin

Cisco Employee

Re: Questions on ACE operation

1/ depending on how you captured the trace (ie vlan capture) you may see the same packet multiple time in trace which confuses traffic analyser.

2/ the standby will have probes to the real server as well. So, it is normal to see probes from the standby

3/ the default pass count is 3.

So it takes 3 successful probe before considering a server as alive.