I am seeking advice/guidance from those much more knowledgeable than I. Please forgive the long post.
I have a couple of sites that have current SME Servers installed. Both have RAID1 in operation. Both also have daily backups done using Darrell’s DAR2 contrib, backing up to USB-attached hard drives.
A couple of weeks ago the backup on Site-1 failed with an email message from the DAR2 script:
“Connecting to /media/usbdisk
mount: no such partition found”
This is a Maxtor 160GB PATA drive in an old ADS Technologies USB X-804-EF enclosure. I went to the site and took the drive back to my office to take a closer look. I connected the drive and tried ‘sfdisk -s’ and the access light illuminated on the drive enclosure and if you listen closely you can hear seek activity; but nothing happened. It eventually timed out. I then tried ‘fdisk’ with the same result - it would not respond at all!
Earlier this week I received the same message from Site-2. Since it is not nearby, I have not retrieved the drive. But sfdisk and fdisk both yeild the same results as for the Site-1 drive. This time it is a Maxtor 300GB SATA drive in a Vantec NexStar 3 SATA to USB 2.0 / eSATA Hard Drive Enclosure.
The timing of the failures (within two weeks of each other) could be pure coincidence.
In the past I have had no problems with Maxtor drives.
A bit more history might be in order. The Site-1 drive is a couple of years old and was a prime drive in a SME Server test machine. It went into the enclosure for use as a backup drive in Site-1 about 4 months ago. It sits attached to the server and turned on all the time.
The Site-2 drive is one of a set of three. Each Monday, the local user swaps the current one out and puts a fresh one in its place. The one just removed is cycled off-site. Each of the three is attached and turned on for a week and spends two weeks elsewhere. This process has been followed regularly for 10 months.
Both units are powered from a UPS, so the power should be pretty clean.
One concern I have heard about with some of these drive enclosures is heat dissipation. I wonder if overheating is the source of the problem.
So this leads to my questions:
1. Does anyone have any other ideas about the possible source of these drive failures?
2. Is there any way to monitor heat levels in USB attached hard drives?
3. Is there any way to specify that the drives should spin down when not in use for a period of time and turn back on when needed? (I don’t want to do that for all of the drives, just the ones that hold the backups.)
4. What suggestions do you have for avoiding the problem in future?
Thanks for your help.
John
3 Replies
2293 Views