For those of you interested, here’s a brief re-cap of the issues that we’ve been having with the site, and the failed attempt to fix it. Unfortunately, I’m just not a high-tech mechanic, so I’m at the mercy of those that supposedly are, although I’m beginning to have some serious doubts about their abilities.
For the first two years of CC, the site used a typical “unlimited” and very cheap hosting service (Blue Host). That means sharing a server with a number of other sites. CC worked great for a while, but after 6-9 months, I started noticing slowdowns and even time-outs (Error messages). Slowdowns are always felt more severely in the back end, like creating new posts and such. This is very frustrating, since keeping the monster fed with fresh content is hard enough, but when there are constant slowdowns and halts, it can be deeply frustrating. And sometimes, large chunks of work would disappear into the ether.
Blue Host claimed that our site was being “throttled” at times, because another site on the same server sometimes had huge spikes in visits. But I had my doubts, because it seemed to happen regularly, and at certain times of the day.
One obvious area of concern was the fact that we host a huge number of images, and rather large ones at that. We made changes to limit maximum images size, but that really didn’t make any real difference. So last spring, at the recommendation of blue Host, I sprung for a Virtual Private Server (“VPS”), at about a 10x increase in cost. After the migration, the site ran super fast and crisply…for a while. Within a month or so, slowdowns started happening again.
Our tech volunteer, Ed Stembridge, tried to pin down the issues, which are the result of “server spikes/overloads”, the results of the server having too many requests at any given time. If this was strictly the result of increased traffic, that would be a great problem to have. And our traffic has increased steadily, to the point where we have been averaging close to 20,000 Page Views per day. But that really doesn’t explain it, as the raw capacity to deal with that is theoretically in place.
Blue Host diagnosed that PHP errors are the root cause of the server spikes. And after some attempts by Ed to ferret them out, it was recommended by BH that we ditch our current “theme”, because it was not being fully supported by its developers anymore.
Now I’m not an expert at all, so I rely on my instinct, and I had my doubts about this recommendation, since the site worked so well when we first moved to the VPS. But I took the recommendation, and contracted with our original site builders to rebuild it with the new theme, one that they use regularly. And it gave us the opportunity to clean up the site a bit, and freshen its look some. Not that I really had the time to give that much attention; I’m still trying to finish up the house I’m building, and…….
Anyway, the new theme went live yesterday, and the server spikes were even much worse, making it almost non-functional, especially in the back end. The web designer spent the whole day yesterday trying to figure out why this would be happening, to no avail.
I’m not a happy camper right now, to say the least. I feel like I got “sold” on an overhaul that obviously isn’t working. There is no good reason the site shouldn’t be working hunky-dory. But it’s not. Unless the present mechanic can figure out what the problem that’s causing these PHP errors and server spikes, I guess I’m forced to find a new one. Makes me thankful for how reliable new cars are. Anybody know a really good mechanic?
Update: We obviously went back to the “old site”, which is now running like a champ, like it hasn’t in a very long time. Must be that can of Sea Foam I dumped in it. Seriously, I have no idea why, but I’ll try to see what’s changed on Monday. And we’ll have to see about re-installing the new theme.