SUMMARY dump problem

From: Liesl Andrico (
Date: Fri Apr 09 1993 - 17:57:29 CDT

Well, replies varied. The most prominate was:

Did you check the cable length? Well, of course, I'm no dolt. But,
just to make sure, I went and looked. Sure enough, one of the cables
was thirty times longer than any other cable on the chain. OOPS. I
spent a day finding someone in this town who will make scsi cables.
(you would ask, they would say, scsi, sure, for mac or pc?)

I replaced the cable, and no problems yet. But, judging from the
other answers I got, I am not holding my breath.

Other replies:

bad tape. Well, could be, but it happens with many tapes.

clean the drive. Yes, but I had already done that.

don't worry about the kernel message, it isn't hurting anything.

information about the kernel message. nothing new there either.

put tape drive first in chain. I will reserve this for a later fix
because I don't have the correct cables to do this and it takes
awhile for them to be ordered.

you are running out of tape. That can't be because it happens even
with a bran-spanking new tape.

dump needs a filesystem name and wont work with a mount point name.
I don't think this is true.

different blocking parameters, and commands. All gave the same
error messages when tried.

put a 2 min sleep between dumps. I will try this. Can't hurt.

try dumping by hand. Same errors.
try tar. try using /dev/null. etc. OK so must be in the chain.

So, I vote for cabling stupidity.
Thank you all for suggestions and comments.

those I remembered to write down:

Liesl Andrico System Administrator
ParcPlace Systems Fon: (503) 220-1423
921 SW Washington, Ste 312 Fax: (503) 220-0208
Portland, OR 97205 USA E-mail:

This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:42 CDT