bad db key error caused by scsi bus?

[darrin]

we're having problems with bad database keys in
our 200mb progress v6.3e12 database running symix
3.0r6. we have an altos 15000 server, 2 486/dx/50
cpus, 97 megs ram, altos/sco unix system V, with
an altos eisa sca scsi controller. sometimes we
lose a file system, sometimes fsck detects errors
like the free space allocated is wrong. we fix
them and a week later get them again. it seems to
be on our striped /data drives-filesystem and the
/bi drive-filesystem. we replaced the scsi cable
and made sure there's 1' between each device.
there's 4 1gig hard drives and 1 tape drive on the
scsi chain. we ran diagnostics on it and they all
passed. it seems to be an intermittant error. we
had a similar problem but more severe before and
someone on the net told us to enable termination
power to all drives. we did and it helped alot
but still about once a week on the average we'll
have a problem. these 4 drives are 1996 seagate
scsi-2 drives. any other ideas or suggestions on
jumper setting combinations. Progress said to
disable write cache but the controller and drives
don't allow us to do that.

????

--
troopers home page
http://www.netcom.com/~troopers/
please reply by e-mail