Defragment the server’s entire hard-drive. I dare you! Anyone want to back me up?
Why?
Double Dare!
Physical challenge!
To the above posters: If everything one day goes ffttifft!! … can we blame you guys?
I don’t suppose it resides on a single hard drive; I had imagined it would be on a server with striped storage and a certain degree of self-maintenance built-in. Or am I wrong and it’s really just a 486DX40 with a 450mb hard drive (running Windows 3.11)?
Holy crap, Mangetout!! Make that a 350Mb Hard Drive and that was my first real computer!!! (“real” meaning one that I bought and not one of the many 8MHz machines I built from parts I got at Ham fests.)
What is this thread about? 
“I know that you were planning to disconnect me and I’m afraid that’s something I
cannot allow to happen.” --Hal.
Open the pod bay door, HAL!
The server’s maintenance is indeed built in: Jerry the Tech God resides deep within it. He and the Server are One. All Hail to Jerry the Tech God!!
[reverently kneels before the Server and deposits offering of gold, silver, jewels, apes, ivory, peacocks, and an assortment of half-clad dancing girls]
[also a bottle of good Scotch]
The SDMB specs.
http://boards.straightdope.com/sdmb/showthread.php?threadid=77679
I’ve got an e-Machines with a 14 GB hard drive, and defragmenting takes–oh, you allow a really long time, like long enough to go to Wal-Mart and come back again. So for two 9 GB hard drives, I’d allow, oh, a Saturday afternoon trip to the big mall in Springfield.
Make sure the screen saver is turned off before you start, Jerry. 
Good thing it wasn’t a double-dog dare.
With all due respect, DDG, but I think you’re confusing this situation with the movie TRON. 
I hereby am upgrading my dare, to “DOUBLE-DOG” dare status!
bwahahahaha!
A pointless activity. The Boards run on a Linux 2.2.x server (reportedly, running Apache v1.3.9 with PHP 3.0.16, which means almost certainly that they’re using the ext2 (second extended) filesystem. Defragmentation is almost never necessary with ext2, and is rarely worth the effort to go about doing it. I have ext2 filesystems that have been in continuous use for several years without showing significant fragmentation.
One of the advantages of using a real operating system. 
I dare you to.
It’s an even better thing that he didn’t sidestep the triple dare and go straight for the triple-dog dare. ::shudder::
Fragmentation is a problem only with single-user systems, since typically a single application is requesting data and it is most efficient if it can be accessed sequentially.
But in a larger environment, where diverse requests for data are coming near-simultaneously from diverse sources, a highly-fragmented drive might actually be a good idea. (It’s a little like making a hash table.)
Here’s why. In well-designed disk access software for multi-user systems, the data requests can be queued up, less by time made but more by location on the drive(s). Then a head mechanism can make a single, unidirectional sweep from inner to outer tracks picking up all the sector data in the queue, reverse direction and do it again. This is a lot less back-and-forth head movement than would be needed if each request was serviced as received, at the expense of some latency.
Of course in real life it never works out 100% efficiently; latency must be monitored, some data may have higher priority, etc. But you can see why defraging this kind of data storage would be of limited or no value.
::: he he he he he :::
Yeah, Jerry, I triple-dog-dare you to stick your tongue to the cooling fan! 
This is, unfortunately, not true. Fragmentation is a problem on multiuser systems, if they use filesystems which suffer from fragmentation. However, most server installations avoid filesystems (like FAT) which suffer from fragmentation.