thanks Zach, another improvement that can be made in GUI in addition to others ..

Serhat

I don't think this connection summary screen is reflecting what has been happening. I have disconnected the link (all sessions broken), it still shows connections that started 11 and 6 hours ago.Besides, I am running Netflow in the branch router that shows 64 MB of Notes traffic sent to the branch but connection summary does not have any stats on it. I might report this to TAC.

10.64.191.40:2358 10.64.101.151:42510 waasdcp1 11:19:13 36.5117 KB 22.623 KB 38%

10.64.191.15:32873 10.64.101.151:27010 waasdcp1 6:45:20 28.2832 KB 9.5088 KB 66%

10.64.241.130:2418 10.64.191.21:445 waasdcp1 0:30:17 19.1846 KB 12.1631 KB 37%

10.64.218.91:1602 10.64.191.21:445 waasdcp1 0:21:14 13.6436 KB 8.0693 KB 41%

10.64.102.76:1927 10.64.191.21:445 waasdcp1 0:21:1 26.7871 KB 16.7832 KB 37%

10.64.241.105:2145 10.64.191.21:445 waasdcp1 0:14:22 8.8232 KB 7.4648 KB 15%

For how long were the connections displayed after you "broke the link"? If the WAE interface goes down, the regular TCP keepalive timeouts don't apply, instead a write timeout applies. The write timeout has no set limit, but is rather based on TCP's retransmission timeout that is dynamically adjusted based on link's RTT. We try to send the backlogged data up to 15 times at exponentially backed off intervals with a max limit of 120 seconds, and in the end, we declare a write time out and close the connection with an active reset.

Zach

i have the same problme also when i want to clear staticts of last week and start staticts from new point i tried "clear statics all" and when i went GUI and choose monitoring i found result of last week and last day i want clear this result

The command 'clear stat all' clears the statistics in the WAE CLI, not the Central Manager. The CM is a repository for historical statistics.

The connection listing can't be cleared. It is a list of currently active connections.

Zach

even clearing the "stats all" in CM is not clearing connections stats in the individual devices.As you suggested, this connection listing can not be cleared. The thing I don't understand how come two connections in the display stayed up even after I broke the WAN link.May be the developers can put an option to clear the connection stats in the next releases.It is frustrating for the customer.

Serhat

Did you see my previous question? How long do they stay there? Indefinitely?

Thanks,

Zach

sorry Zach, I missed your question. One particular connection is still up even though the server was moved to another site last night. Even "refresh" is not clearing it.

Serhat

Can you please provide the output from the following command for one of these connections:

sh tfo conn

Thanks,

Zach

Zach,

The output is attached. The "sh tfo conn" display shows different "open duration" than the gui output. I have shown both in the file.

I also had problems clearing out filters once it is in place. I had to do several "reset sort"s to clear the filters.

Serhat

These connections don't look stuck, so much as there is a discrepancy between the start time shown in the CLI and the duration in displayed in the CM.

Do you see this discrepancy for all of the connections?

Can you provide the same output again, but also include the peer WAE and the 'sh clock' output from both WAEs and the CM?

Thanks,

Zach

Zach,

The connection has been dropped so many times but the first two connections stayed in the display. In the central device there are some connections that show 200 hours uptime certainly not correct especially after the connection recycled several times.I will stop spending time on this one, thanks for your support.

Serhat

i think this issue may caused due to system.monotoring.collect rate which remote WAEs send its staticts to centeral manager its default 60 so when connection in remote sites terminated the centeral manager know that after 60 or less (when remote waes send to c.m its staticts)so when you do sh tfo stat (in remote wae) and compare it to c.m it must not be the same