brent saner on 29 Apr 2008 06:46:24 -0700

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] Testing ethernet/cat5 wiring

Hash: RIPEMD160

Brian Vagnoni wrote:
| BODY{font:10pt Tahoma, Verdana, sans-serif;} Loop back connector, 1
| - 3 and 2 - 6 should work. Search for it and you will get a peachy
| cool diagram. I think most have testers here, I know I have 2, plus
| a butt set, and a gen toner combination. They aren't that expensive
| and save time. You could try pinging the interfaces and looking for
| packet loss. Increase packet size to the size of your mtu for stress
| testing assuming this is over a LAN.

noting he said without an expensive testing tool, i think the
specialized tools aren't necessary for initial diagnosis. esp. if it's
something localized to only a couple users.
| Too much broadcast traffic, packet loss, auto negotiation on switch
| ports also another speed killer.
if this were the case, it's most likely other users- himself included-
would also be affected, depending on the network layout. i'm pretty
sure we can rule out software/data-related issues.
| ping
| ping sweep, find out what's on your network.
a ping flood to the specific users' workstations would be more
| You need to do cpu, disk, memory, and network stats on your server
| to find out if the server/s are maxing at any of the previously
| mentioned categories.
again, rather convinced it's not a data/configuration issue.


if you out the jacks on the cables yourself (as i now see you did in a
recent response), this sort of behaviour is almost always a symptom of
a botched crimping/jacking job. no need to feel bad; we all do it from
time to time.

i'd go with a voltage meter as well.

- --
brent saner.
gpg info at
(this is a shorter sig.)
Version: GnuPG v1.4.3 (GNU/Linux)


Philadelphia Linux Users Group         --
Announcements -
General Discussion  --