Yes, the new db server went into production Friday evening and has been in production since. It's handling the workload exceedingly well and has actually made life easier for the webserver. It still might have a configuration issue or two to work out, though.
I've also recently put in a new, vastly improved version of post_info, so I'm hoping we don't see any more of those "deadlock" errors. And this version of post_info should've also contributed strongly to the webserver's easier life.
We haven't put the new webserver into production yet, though. That's the first job for the new programmer, who starts the 25th. It's no longer a crisis-reaction type of project, but the box wasn't cheap, will improve performance quite a bit more, and we've got other plans for the current webserver.
I and my British counterparts all agree that it'd be very desirable for us to get the behemoth Dell PE4600's out of the rack. They consume a ton of space and energy and can be used at the office as development machines. One of them (iHub's old db server) is just sitting idle now and the other (SI's db server) is way more box than was needed for that job.
I've finally decided against moving SI's db onto iHub's new db box. The initial plan was to do that, then mirror the db's onto a hot-spare box for redundancy and quick fail-over coverage. Instead, I'll be replacing SI's db with a new box like iHub's new one, then mirroring them both onto each other, knowing that either box can handle the workload of both sites in an emergency. Though the mirroring itself will add an unknown amount of workload to the machines. Shouldn't be much.
