Wireshark-users: Re: [Wireshark-users] Will EMF interference show up in Wireshark?
On Thu, Mar 10, 2011, at 16:10, Larry Dieterich wrote:
I'm troubleshooting a crashing Helix database server hosting 20-25
clients on a switched network.
The network is a combination of Wired and Wireless clients.
This an industrial setting with numerous motors and electric ovens
and other potential source of EMF.
I am sniffing the wire connection to the server, looking for network-
related causes of the crashes.
Note that only the Helix server software crashes, not the OS on the
host hardware. All other functions provided by the server host
machine seem normal at all times.
Currently, the Helix server crashes show on the wire as a pause of
normal Helix I/O traffic, followed by FIN ACK packets sent from the
server to the clients, this is followed by ACK packets from the
clients to the server.
I'm trying to rule out EMF factors in the problem and I'm hoping
that someone on this list has experience with computer or network
problems caused my EMF.
My questions-
Would EMF interference manifest on the wire? If so, what sorts of
packets or patterns should I be looking for in Wireshark?
Question is whether you'd even recognize it as due to EMF and not,
say, a flakey NIC; or, broken wiring inside a still-intact cable;
etc. Best way to tell whether EMF (or broken wiring) is an issue is
to plug in, say, a Fluke DSP-4xxx/OmniScanner/DTX level meter or the
equivalent and test the connection from end to end during the day when
everything is running (i.e., with the motors, ovens, etc., running).
Needless to say, they don't come cheap. But you can rent them. Or
pay someone with them to do the job for you and provide you with a
hardcopy/PDF report of their findings.
Does it seem unlikely that a problem related to EMF would only
affect one software process on one computer?
It might or might not. Depends how strong the EMF is to the one
computer it's crashing on; what's insulating the said computer, etc.
Remember the Inverse Square Law. However, it does sound rather far-
fetched without knowing more details. Of course, if your Helix server
is sitting in the same room as one of the Large Hadron Collider's
superconducting magnets at CERN, then yes! ;-)
--
Reality Artisans, Inc. # Network Wrangling and Delousing
P.O. Box 565, Gracie Station # Apple Certified Consultant
New York, NY 10028-0019 # Apple Consultants Network member
<http://www.realityartisans.com> # Apple Developer Connection member
(212) 369-4876 (Voice) # My PGP public key can be found
at <https://keyserver.pgp.com>
Attachment:
smime.p7s
Description: S/MIME cryptographic signature