I'd say carla did a real descend job here:
I started naming stuff as "to intuitive for me" a while ago.
Ummm ... does "THAT'S IT" mean that they run in non split tunnel mode,
so if you go to the Internet it first routes it through their network?
Or do they just shut off Internet access altogether when you're
connected to the VPN?
No split tunnel, everything goes through the vpn.
I haven't tried setting a route myself though. I find vmware much more fun. I can just close the window and work goes away.
the machine froze on me when I attempted to suspend a vm.
So I shut it off. restarted and the raid software decided the raid array wasn't clean and is now resyncing it.
all 600 gig of it.
I wonder how it knows which one is good.
That said, the machine is still useabe, if slow because of all the disk io.
heh, I should have installed that this morning.
You do know that you can cat /proc/mdstat to monitor the recovery
yeah, that's where I got the 66 hours from.
I dunno what happened, but after about a half hour, the disk went mostly idle and the syncing went a lot lot faster, only took 3 hours total.
Wish I had run iotop!
It turned out not to be my vm's that were holding the cd rom drive, but hddtemp.
Today I learned what fuser is. :-)
A real sysadmin I will never make.
they've introduced gpxe, a pxe replacement, which also knows how to do DNS, HTTP and FTP instead of just the good old TFTP...
boot.kernel.org offers images to be downloaded and bootet instantly.
PXELINUX is the right way to go. It lets you use stock PXE boot ROM's, and then it can chain load to whatever you want to do afterwards.
I recently moved my home server to new hardware. It's got a different SCSI chipset than the old hardware (Symbios Logic 53c1030 PCI-X Fusion-MPT Dual Ultra320 SCSI), but I took care of adding in the correct modules to the kernel before moving the hard disk over.
My tape drive isn't functioning properly. But it isn't throwing any errors!
There are no SCSI errors reported. dmesg says nothing of interest. Backups appear to be completing, but during the verify phase of my backup script, gunzip reports that it is receving a corrupted data stream.
What would cause the SCSI or SCSI Tape layers to deliver corrupt data without reporting an error?
methinks, most tapes are writeonly anyway.
Have you tried using a fresher tape? Is possible the medium itself is the problem?
What would cause the SCSI or SCSI Tape layers to deliver corrupt data
without reporting an error?
old crappy tape?
Yup ... tried it with several different tapes.
It's good to know we all think alike after all this time.
Not a single outside-the-box-thinker among us. :-)
While it is pretty damning that it worked before and not after the hardware change I'd guess that if none of the software is throwing any error information, then nothing is wrong there and something is wrong with the tape drive.
Is it possible that something got zotzed by the new hardwhere when it was hooked up?
Can you borrow another tape drive and make sure it's not that?