On Sun, 2005-01-02 at 07:55 -0500, Marc Sherman wrote:
> Andrew Bartlett wrote:
> Can you recommend any network diagnostics I can run on the win2k client?
Take a network sniff, using a tool like ethereal.
> Looking into it further, it seems to be specifically "breaks on first
> access". It see
Andrew Bartlett wrote:
As a rule, we don't rule these things out that easily. I would properly
investigate the network layer, just in case. CIFS is unique in the way
it holds open long sessions which can be rather sensitive to
disconnection (with subsequent failures in oplock breaks, and other
th
On Sat, 2005-01-01 at 21:52 -0500, Marc Sherman wrote:
> Michael Lueck wrote:
> > Typically connection resets are a hardware based problem - LAN cable
> > first, the NIC or hub/switch.
>
> I'd be very surprised if it's hardware -- the samba box hosts a lot of
> other services used by the windows
Michael Lueck wrote:
Typically connection resets are a hardware based problem - LAN cable
first, the NIC or hub/switch.
I'd be very surprised if it's hardware -- the samba box hosts a lot of
other services used by the windows boxes, including email (smtp and
imaps), which get hammered a lot heav
First I did a complete workstation reload which utilizes DOS and Win2K. I got
some connection resets in the log.
Then I copied 4.9GB of files from the server down to the workstation, no resets
in the logs.
Thus, possibly the DOS client caused some resets, or maybe IPL's of the
workstation, but n
Typically connection resets are a hardware based problem - LAN cable first, the
NIC or hub/switch.
I just happen to use Debian Sarge here, and have a PDC running that level of Samba, AND Win2K SP4 where I just got around to those M$ hot fixes. Just for you I'll pull a few hundred MB over the wire