bad cable story ... but why ??

[someone_4]

We had a server go down yesterday at around 5pm. The issue was we could not ping , telnet, or bring things box online. It is an A-class.In trouble shooting this problem we took this cable from the A-class and plugged into a D class that is right next to it. And the cable worked fine. So this would lead us to think that the cable was good. So we put a call to HP they send a tech out , replaced the base unit. Swapped the network card , ram and hard drive. And this server still didnt work.
Now 3am ... someone decides to try a differnt cable but being that was one of the first things that we tested we thought that would not matter. But we tried anyways. After replacing the cable the server came online.
My question is would there be any logical reason why the cable would work on the D class but not in the A class?

Thanks

Richard

[Patrick Wallek]

When you say cable, I assume you mean a CAT5 network cable? You may have just had a bit of bad luck when you switched the cable to the D box. If there was a short in the cable it may have made the connection correctly just for the little bit that it was plugged into the D. Maybe you held your mouth just right! :)

Cables can be funny things. I think I'd chalk it up to bad luck and make sure that the bad cable gets disposed of.

[A. Clay Stephenson]

Hi Richard,

I can think of two reasons immediately. If the D-box is running 10BaseT and the A-box is running 100BaseT - your cable may not be certified for CAT5. It's also possible that the mechanical connection on the D-box is better and the D-box is thus more tolerant of a marginally fitting cable.

Regards, Clay

[someone_4]

whats weird here is this cable has been working forever and then all the sudden it went out ..
weird huh ?

[John Bolene]

You must have been using this server for years with lots of network traffic and just plumb used up all the electrons that were in the copper wire.

When you plugged it into the D, it partially recharged a few electrons (since it is a much more powerful server) and it was able to work for a few more minutes.

You then plugged it back into the A and since it does not have the power that the D does, it could not push electrons thru the depleted cable.

This message went thru a time warp, it was actually posted on April 1.

[MANOJ SRIVASTAVA]

Hi Richard

I agree with Clay , D calss server has a default 10 Base T port , and if you have any probs with connectivity on A class server the better way is to swap the cable since the results on using the same cable for a lower speed port may be errorneous. Since u will be using a 100 MBPS port for real work .

Manoj Srivastava

[Patrick Wallek]

"All of a sudden" tends to be how cables go bad, for what reason, who knows.

I had added some disk to a system a couple of years ago. Got the SCSI cables attached, the system saw all the disk, VGs and LVs created, no problem. I called a co-worker and he started copying data. A few hours later I get a call that the copy stopped. I check the logs and the disk drives, etc. everything was good. Switched to a different SCSI cable and all was good with the world again. Why it worked for a few hours and then died....who knows!

[melvyn burnard]

For many years now since i started working with computers, I have had a little rule htat says: 80% of hardware problems end up being a cable somewhere :-]
Some of the other postings say it all. The cable may have had a latent defect, and got nudged/moved which "killed" it. moving it again corredcted the symptom, but I would not trust that cable anymore.

you win some, you lose some.

[Sachin Patel]

Hi you not the only one.
Last week I spend 8 hour from 10 evening to 6 in morning (I was on-call). Problem was scsi cable.
We have tape drive hook up with K-class. I can see the tape drive I can run mt command. I can run ioscan and none of them give me error. I start changing scsi cable from system to tape. As soon as I change last-last-last scsi cable on chain. Server was working again.
Long night.

Sachin

[Bill McNAMARA_1 · Accepted solution]

It's always the cable!
and it's such a shame it doesn't show up on ioscan! I'd bet the problem is more mechanical, and related to the spring contact being tighter on the D..

I've had strange experiences with fibre too.. on a D with a weak laser a short cable didn't transmit light well but a long one did? It was the complete reverse on a K class with identical cables.

Later,
Bill