"Single request" in Wingate sometime hangs

[olaf.krause]

After I updated to 5.2 I have sometimes the problem that one single request in wingate hangs. That means that not the complete app. hangs but this specific URL or POP3 request never finishes or times out. If I close the client applications (browser, Email) everything works OK also opening other applications in parallel is working OK.
The problem with that "single request hangup" is that I'm using a dialup connection for which I have to pay for based on the time it is up. But this connection is never closed if the described error occurs!

regards

Olaf

[adrien]

so this happens with HTTP and POP3? Or just one protocol?

Adrien

[olaf.krause]

With both protocolls; it happens approx. once a day.

regards

Olaf

[olaf.krause]

It seems to happen more often with the POP3 protocol (third time today). Could I switch on some debugs to help you with the analysis?

ciao
Olaf

[olaf.krause]

OK here a part of the POP3 logfile when it happened:
(removed date and replaced IP to get shorter lines). The request 17:23:32 was hanging and I terminated it at 17:30:32. As you see at 17:30:46 the normal request time would have been 33s but I had to terminate it after 421s.
regards

Olaf

---- log ----
17:23:32 IP Guest 0000002983 Requested: POP3: x at xyz.xyz.de
17:30:32 IP Guest 0000002983 Traffic 427 70 55 3419 421s
17:30:33 IP Guest 0000003032 Requested: POP3: y at xyz.xyz.de
17:30:33 IP Guest 0000003032 Traffic 63 62 47 47 1s
17:30:36 IP olaf.krause 0000003035 Requested: POP3: z at xyz.xyz.de
17:30:36 IP olaf.krause 0000003035 Traffic 167 61 42 225 1s

17:30:46 IP olaf.krause 0000003036 Requested: POP3: x at xyz.xyz.de


17:31:18 IP olaf.krause 0000003036 Traffic 135797 196 181 135844 33s
17:31:19 IP olaf.krause 0000003038 Requested: POP3: x at xyz.xyz.de
17:31:19 IP olaf.krause 0000003038 Traffic 63 62 47 47 1s
17:31:22 IP olaf.krause 0000003042 Requested: POP3:z at xyz.xyz.de
17:31:22 IP olaf.krause 0000003042 Traffic 167 61 42 225 1s

[olaf.krause]

Any ideas - or do I have to switch back to 5.1?

[olaf.krause]

It's 99,9% a 5.2 problem. After a went back to 5.1 everything works ok (for 3 days now).
Seems that you "oversychronized" your threads?

ciao

Olaf

[adrien]

Was this with data scanning (i.e. AV or content filtering)?

We changed the way we drip-feed data to email and web clients to stop them timing out whilst large files were being scanned on the server.

Adrien

[olaf.krause]

Both (5.1&5.2) with Visnetic AV Plugin (newest avail. version).
Olaf