|
As I mentioned, I work with computers for a living, though as a software engineer, not a sysadmin like Skinner or EarlG. Let me give you some facts.
Computers go down. It happens quite often in the business - maybe it's a flaky memory module in a server, or that last software update to the PHP code caused a crash, or that zillionth post filled up the hard disk and caused the database to freak out. There's bunches of things that can go wrong with a web site, especially one like DU, which has close to 100,000 members, and untold millions of posts. To handle that demand, you need quite a bit of hardware - multiple web servers, some of them handling the databases, others running web servers, others that handle tasks such as talking to advertisers like Google and serving up ads to help pay the bills. You've got dozens of cables, you've got routers and other network hardware, and all of these pieces do require maintenance. There are ways to minimize the single points of failures - using RAID arrays instead of single hard disks, using multiple web servers and database servers so if one goes down,
There are hackers and such out there, and some of them may be freepers looking to do a number on DU, but if it was a hacker attack, there would be definite signs. This particular bout of DU flakiness seems to be caused by failing hardware, and tracking down exactly which of the multitudes of servers is failing is tricky. It would be easier if the admins could bring down DU for a day or two - then they can systematically test each server with memtest & other tools and find the problem, but that would cause a lot of problems, so they have to do what they can while keeping everything running live.
|