Outlook Express client can't download mail from mail server (POP3 protocol WORKS, I can telnet, it authenticate user, collect number of messages, start to download and TIME'S OUT)
Mail server is not an issue (from other test clients mail is normaly downloaded). Type of mail server is not an issue (tried with sendmail, exchange, mdaemon - same). Mail client is not an issue (outlook, express, eudora, pocomail)
If there is no e-mails on server everything is fine. If there is a single mail on server, client times out (after successful authentication).
It seams to me there is Cisco - Matrix - ISA - POP3 issue. Any ideas?
Mail server logs:
Thu 2003-03-20 17:09:33: 616:6135 Accepting POP connection from xxx.xxx.xxx.xxx Thu 2003-03-20 17:09:33: 16:6135 -- +OK xxx.com POP MDaemon 6.7.2 ready MDAEMON-F200303201709.AA0933331MD1025@xxx.com Thu 2003-03-20 17:09:33: 616:6135 -- USER email@example.com Thu 2003-03-20 17:09:33: 616:6135 -- +OK firstname.lastname@example.org... Recipient ok Thu 2003-03-20 17:09:33: 616:6135 -- PASS ****** Thu 2003-03-20 17:09:33: 616:6135 -- +OK email@example.com's mailbox has 1 total messages (2254 octets). Thu 2003-03-20 17:09:35: 616:6135 -- STAT Thu 2003-03-20 17:09:35: 616:6135 -- +OK 1 2254 Thu 2003-03-20 17:09:36: 616:6135 -- LIST Thu 2003-03-20 17:09:36: 616:6135 Sending LIST response (not logged)... Thu 2003-03-20 17:09:39: 616:6135 -- RETR 1 Thu 2003-03-20 17:09:39: 616:6135 -- +OK 2254 octets Thu 2003-03-20 17:09:39: 616:6135 Sending C:\MDAEMON\Users\xxx.com\mimi\md50000000010.msg to 126.96.36.199 Thu 2003-03-20 17:09:39: 616:6135 Transfer Complete. Thu 2003-03-20 17:11:23: 616:6135 Winsock Error 10053 Software caused a connection abort. Thu 2003-03-20 17:11:23: 616:6135 POP session abnormally terminated, 2259 bytes transferred!
Winsock Error 10053 Software caused a connection abort means "An established connection was aborted by the software in your host computer, possibly due to a data transmission time-out or protocol error". So, this sounds like transmission problems on the network.
I would start to check out the adapters, switches and cabling. Try to run some transmission tests from different clients to different servers to find out in which area the problem is situated. I use very often a simple FTP for that and try to put and get a big file.
everything goes trough, except messagess. This is quite large enviroment (800 workstations, over 40 servers). Linux box (NAT, squid) was in place instead ISA server. When we cold-swap hd's (same hardware, cables, switches) everything works perfect. I tried almost everything (as you can see I installed two e-mail servers and was runing trough building like nuts, pluging it everywhere).
I plan tomorow to: 1. replace hardware, cables, and switch 2. try pure RRAS NAT, Wingate, WinRoute 2. reinstall ISA 3. whatever anyone recommend 4. whatever anyone recommend 5. whatever anyone recommend 6. ....
Never seen something like this (have over 120 good ISA installations behind me)
Maybe more information will help: workstation [192.168.3.100] -ethernet- [192.168.3.1] cisco router [192.168.4.101] -leased line 128kbps- [192.168.4.102] cisco router [192.168.19.225] -fiber- [192.168.19.170] matrix switch with layer 3 routing [192.168.20.1] -ethernet- [192.168.20.200] ISA [xxx.xxx.xxx.xxx] -ethernet- [xxx.xxx.xxx.xxx+1] e-mail server
1. replaced hardware (brand new 330 eseries), cables (shielded & tested with Fluke), switch (SuperStackIII) and router (Cisco with latest sw) 2. tried pure RRAS NAT, Wingate, WinRoute (winroute crashed cos too many connections) 3. reinstalled ISA (twice) 4. installed Wingate on Win98, Win XP
POP3 timeout after authentication!
installed SUSE 7 with ipchains and it works installed RedHat with ipchains and it works
anyone have same expirience? This seams to be Microsoft TCP/IP problem!
hmm... so changing network related components doesn't change anything.
Does you see that behaviour also when you test from a client directly connected to the hub on the ISA internal interface?
Have you already taken a network monitor trace on the ISA internal and external interface to see what is really happening?
Have you ran already some FTP tests (try to put and get a big file) from different clients to different servers to find out in which area the problem is situated?
Have you verified that the adapter settings 10/100 Mbps and Half/Full Duplex are correctly set? Keep in mind that the AutoDetect feature on adapters and switches not always work as they should. I recommend always to configure servers with fixed settings (no AutoDetect).
However today other dpt tried to update firmware on one of the routers (Matrix) and they found that they can't do anything (router locks when they try to enable and does not allow conections on serial!). And some other strange things started to happen.... etc.
to make long story short, we are now waiting for new router....
Have you already tried to disconnect the Matrix router from the hub closest to the ISA server and then test from a workstation directly connected to that hub? You should physically partition the network in order to determine the cause of the problem.
Hi, What was the solution to this problem? I'm having the same problem and I can't find a solution. The client can't download the mail from a specific adsl line, from another connection it works. It logins and find messages, but when it tries to download it stops and give a time out error. I don't have access to mail server logs.