On Fri, 18 Jun 2004, Bruce Momjian wrote:
> Marc G. Fournier wrote:
>>>> bandwidth related, I'm suspecting that alot of the problem is the Linksys
>>>> switch we threw in when we first setup the servers ... at the time, I
>>>> didn't know that it was a bad idea :( If you do a trace to the servers,
>>>> you'll see a packet loss between 200.46.204.1 and the server itself that
>>>> just *shouldn't* be there. Apparently, on an unmanaged switch, if you try
>>>> and hard code the ethernet on the server to 100baseT/full duplex, the
>>>> switch, for some *stupid* reason, will downgrade itself to half duplex, so
>>>> you get a whack of errors ... right now, as a result, we're running
>>>> half-duplex on the server(s), which, of course, results in high collisions
>>>> *sigh* We're ordering in a new managed Dell Procurve switch next week, to
>>>> get that issue fixed ...
>>>>
>>>> As to the hardware support itself ... see my other email ...
>>>
>>> Marc, I think you need voodoo dolls on your computers or some exorcism
>>> ritual. Come to think of it, the problems mostly started when moving
>>> the servers to Central America. Hmmm. :-)
>>
>> Oh, what a man of short-lived memory ... do you not recall the fun while
>> we were at Rackspace? :(
>>
>
> Oh, yea. So your hardware bad luck isn't based on location. :-)
Exactly :) At least where we are now, its easier for me to fix those
problems, just wish I could do it faster, that's all :)
----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664