Now all I need to do is convince you to take serious the idea of a 2500-or-so thread cap and split and lock the older parts of the 20-30 or so largest threads. The best part is? If it doesn't actually dramatically make an impact on server performance, you can merge the threads back together. But if it does help? It makes a great case for locking posts over a year or two old for helping make the best out of server performance and minimize the addition of any new threads that keep having to be pulled from the hard drive. I suspect, however, it's going to make a far greater positive performance impact than you think.
Answered here:
Regarding thread cap/splitting/etc... Please take note that the cap on threads they use in your link is 30k and 25k posts. That would be (at our 50 posts/page) 500-600 page threads. That's where they set their cap. I would be okay with that. We aren't anywhere near that. We have a couple dozen threads that hit triple digits (aka 5k posts, a fraction of their cap). Clearly the forum software can handle long threads, if their recommended cap to help performance is like 5x the length that our longer threads hit.
Why would we set a 50 page cap when their cap size is 500-600 pages? Clearly the forum software can handle that, if the hardware is sufficient.
Your only response was pointing out what hardware they used to achieve this.
Yes. I know. That's why I think we need a server upgrade. From 10 seconds of Googling, it looks like more powerful than that is available on a dedicated server for <$100/mo.
Archiving old threads and keeping people from dredging them back up isn't a bad thing, though... especially given how most necroposters will reply to years old dead and resolved threads, ignoring the post date, thinking the thread is still relevant and active and completely ignoring the "this post is old" warning. Then suddenly, BAM! Several newer forum people are posting answers to four year old discussions, and the database is having to pull from the hard drive for yet another active thread.
This is a feature, not a bug.
People restarting old conversations is a good thing.
People having new thoughts, followups, etc.
It's great when people reply to years old threads with an update, or new data, a new study, whatever. If they had to start a new thread, it would lose all the context of the old conversation, and all the people who participated in the conversation would likely never see it, versus if it shows up in their replies and they can engage again.
You've described the major DOWNSIDE of locking years old threads--people can no longer reply to them.
These threads are in the database already, they're being accessed... why would locking them help? Well, only because it means they'll be accessed less (because people can't reply). That's not a benefit. We
want people accessing threads. We
want people replying to threads.
We could drastically limit the database size if we only allowed a cap of 1 reply, and then autolock. Very little access, very little interaction. Very unhelpful to actual conversations.
The idea to lock threads is antithetical to forum conversation and usefulness, and the fact that them being unlocked means they might get a reply, which gives them more traffic, and more replies, is, well, the point of having a forum.
Saying "well hardware can easily support 500+ page threads, but we'll cap ours at 50 pages because we're going to cheap out on it" isn't the solution. Remove dead accounts? Fine.
Restrict active users ability to participate and enjoy the forums? No. Please, no.