Solving a Dell/VERITAS Backup Problem

We have several Dell PowerEdge 2550 servers, all of which are running Windows 2000 Service Pack 3 (SP3) and various applications, including Microsoft Exchange 2000 Server. The problem is that when I use VERITAS Software's Backup Exec to back up our main file server, which isn't the server that runs Backup Exec and which I back up nightly to a Spectra Logic tape library, the server reboots in the middle of the backup. The reboot is a total reset of the system. The machine exhibits no other problems and reboots in this way only during backups. I upgraded the Backup Exec remote agents on each system to the most recent version and even upgraded from Backup Exec 8.6 to Backup Exec 9.0. As far as the file server goes, I've done everything I can think of: I've upgraded the BIOS, removed the local agent software, reinstalled Win2K, and replaced almost every piece of hardware on the system. Nothing is working. What more can I do?

You're in luck. I've had a great deal of experience with Dell servers over the years and have run into this obscure problem (i.e., servers that inexplicably reboot during backups). Dell's support forums contain occasional references to the problem, which seems to occur only in a specific (yet popular) system configuration: Dell PowerEdge 2550 systems that have particular motherboard and firmware revisions and that run Win2K and (from what I can tell) any recent version of Backup Exec. As you've witnessed, servers with this configuration spontaneously reboot when being backed up directly or remotely. Other backup applications typically work without incident. I've no idea what causes the problem, but I know the solution: Ask Dell to swap the server's motherboard with a newer revision that resolves the conflict.

Many Dell support technicians appear unaware of this problem, and I needed to report several incident escalations before convincing the involved technician to research the incident sufficiently to find the internal Dell document that acknowledges the problem. My persistence paid off, though, and eventually Dell issued a motherboard replacement for the affected system. (The company didn't share any specifics about the fixed or offending motherboard revision numbers, but Dell is aware of the problem.)

After the motherboard revision, the spontaneous reboots went away—permanently. Interestingly, I also noticed that the system returned with a more recent version of the BIOS—A07—which wasn't available during my earlier attempts to solve the problem on my own. Although a simple BIOS update (to A07 or later) isn't likely to fix the problem on its own (according to the clues I've garnered to date), I recommend trying such an update before you call Dell—just in case.