Wireshark-users: [Wireshark-users] How to keep Windows wireshark(x86_64) version from crashing wi

Date: Tue, 14 Jun 2011 19:07:59 -0700


(resending to correct list (funny how sending posts to 'request'
doesn't work...*doi!*)....

I have upgraded to latest 1.6.0

Can't get it not to crash when monitoring my internal GB net (with no filters on).

If I turn on filters to exclude all SMB traffic, OR if I monitor my 2nd interface that gets
little traffic...it works fine...but
when monitoring a full speed dump....crashes so fast....it's not funny!

I've tried setting up ring buffered capture files 4GB each, ringsize=8,

(no go).

Single file was nogo.

No capture file was 'nogo' (but that goes to page? I only have 24G phys and about 48G of page file on top of that). It's a 6-core machine, with 3.33GHz Xeon, so it's no slouch, but it is running Win7-64. It is almost always the case that I can't stop a trace before it crashes....a debug window comes up with some error or another.
This time saw several messages all in a row, that said:
------------------

18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
18:57:39          Warn Too many taps queued
---
Then an error popup from:
[Microsoft Visual C++ Runtime Library]
The application has request the Runtime to terminate it in an
unusual way.
Please contact the application's support team for more information.
-----



I've had this problem for a while, but worked around it with filters....but I want to be able to trace loaded SMB2 traffic
that's is being read from the server at up to 112-115MB/s,
though 80-90MB/s is more typical...

Ideas?