I'll get you the trace file later today. From memory, I believe the last ACKed packet
was about 732 bytes in size, with most of the previous ones being 1432 bytes. After
this packet is ACKed, the server keeps sending more data in mostly 1432-byte packets,
while the client keeps ACKing the same old packet.
When the server reaches a number of unACKed packets equal to about half its window
size (23K window) it finally starts retransmitting as well.
I don't have the pcap file handy or I'd attach it, but I am curious to know what may
be brewing in your noggin.
The part that freaks me out is the microscopically short delay between client
retransmissions. It's almost like the stack or socket dll is itself misbehaving.
-----Original Message-----
From: James Washer [SMTP:[EMAIL PROTECTED]
Sent: Thursday, September 04, 2003 6:32 PM
To: [EMAIL PROTECTED]
Subject: Re: [RLUG] [OT] TCP/IP Weirdness
what sequence number is being acked ( in the repeated acks ) , what it the
sequence number and size of that 'last' packet, and what get's acked after the client
receives that last packet....
mail me the pcap file if you care to.
- jim
On Thu, 4 Sep 2003 17:55:01 -0700
"Robinson, Eric R." <[EMAIL PROTECTED]> wrote:
> This is OT because none of the participating equipment runs Linux, but I
> figured it was worth asking anyway since some of you are network geeks.
>
> I have an application called ProLaw that runs off a NetWare server on the
> other side of a PIX firewall. (The client is inside, the server is outside.)
>
> Most of my Win2K workstations perform fine. But two or three of them slow
> way down after a few hours of use, to the point that queries that normally
> take 10-15 seconds take 1.5 minutes! Meanwhile, other clients on the same
> switch keep working fine.
>
> Ethereal traces show that during the slow periods, the client retransmits a
> lot of ACKs of the same data. It will transmit the same ACK maybe 9 or 10
> times in a row with only about 0.000002 seconds between. Then the server
> finally retransmits a packet of its own.
>
> Finally, the client ACKs all of the previously received data at once and
> catches up (about half of its window size). It does this over and over
> again, slowing down the overall communication.
>
> Can somebody please offer some insight on what could be happening here?
>
> --Eric
>
>
> _______________________________________________
> RLUG mailing list
> [EMAIL PROTECTED]
> http://www.rlug.org/mailman/listinfo/rlug
>
_______________________________________________
RLUG mailing list
[EMAIL PROTECTED]
http://www.rlug.org/mailman/listinfo/rlug
_______________________________________________
RLUG mailing list
[EMAIL PROTECTED]
http://www.rlug.org/mailman/listinfo/rlug
