For a network administrator of a small company, troubleshooting takes on a heightened level of anxiety when the equipment that needs the attention is the primary file server. You can imagine how Mike felt when one day his file server crashed without any warning signs and would not recover with a reboot. In this week’s From the Trenches, we’ll follow Mike through this crisis and see how he used multiple operating systems to rescue the valuable data on that server.
Get insights From the Trenches
You can learn quite a bit by reading about the methods other administrators and engineers use to resolve challenging technology issues. Our hope is that this column will provide you with unique solutions and valuable techniques that can help you become a better IT professional.
If you have an experience that would be a good candidate for a future From the Trenches column, please e-mail us. All administrators and their companies remain anonymous in this column so that no sensitive company or network information is revealed.
Mike works for a third-party administrator of health care claims. The operation is fairly data-intensive, and some of their most important data shared space on the Microsoft Windows NT 4.0 server that crashed. The day that the server gave out, there were no warnings of the failure and nothing happened at the moment of failure anywhere else on the network.
After some initial troubleshooting, Mike swapped out the CPU, RAM, and SCSI controller in the server to check for a hardware failure. When that produced no results, Mike suspected that the motherboard had something to do with the server’s failure. When he got another motherboard from his vendor, the server crashed again.
In the meantime, he had been using the hard disk from the original server as an alternate drive (and not the boot disk) in a substitute machine. When he got another new motherboard, he popped the original hard drive into the rebuilt server. The rebuilt server also crashed, and this time it refused to reboot into NT at all.
“At this point, I was right on the verge of tearing the remainder of my hair out when I discovered that the hard drive’s aluminum housing showed signs of having been damaged in handling, apparently dating back to the original server days,” Mike said.
Convinced that the hard drive had been the problem all along, he decided to replace it. Getting to the data he needed on the faulty hard drive now became the concern.
Mike bought two new Ultra ATA 20-GB drives. On one of the drives, he installed Red Hat Linux 6.2 in a FAT 16 partition and formatted a second partition on the same drive for FAT 16. He installed a module on the Linux machine that would recognize NTFS drives, so that he would be able to transfer the recovered files to his new NT box, which would be formatted in NTFS.
“I couldn’t trust the motherboard in the original server anymore, so I installed the faulty drive on another NT box. I ran Samba on the Linux box and connected across the network and transferred the files from the bad drive to the FAT partition on the Linux machine. It was literally drag and drop,” Mike said.
Now he took his other new hard drive and put it in the original server, formatting it for NTFS and setting up his new NT 4.0 server on this machine. After testing the new server and making sure it functioned properly, Mike transferred all the directories and subdirectories from the FAT 16 partition on the Linux box onto the new NT server. The valuable data was saved.
Living happily ever after
Everything is up and running now at Mike’s shop, and the motherboard he thought was having trouble has been performing flawlessly.
He suspects something got corrupted in the boot sector of the failed disk drive, but he has now put it in a Linux test machine, and it has run there without failure since. He still doesn’t trust this hard drive enough to put any valuable data on it, but it works well on his test machine and with the Linux operating system.
“I don’t proclaim Linux to be the ‘do all, fix all,’ but my opinion on it is you have to put the square peg in the square hole and apply Linux to things that it does and does well,” Mike said.
Next week, we’ll again visit with Mike, as he goes dumpster diving and comes up with the components to set up a Linux Sendmail server and firewall to meet the company’s growing needs and the desires of a penny-pinching CEO.
A hard-drive friendly system?
Discussing his experiences with this failed hard drive, Mike and I both had the impression that Linux is very friendly to any hard drive it is asked to recognize. Have you noticed this attribute in Linux or any other operating system? Why do you think it might be? Send us a note or post a comment in the discussion below.