September 7, 2024: Essential Services
Well, heck. I had a long blog post I'd been working on for days, and then yesterday just as I was about to post it... the Wongery site went down. I thought at first maybe it was an issue with my internet connection, but no; I could load other sites just fine. But I also tried going to another site that was under the same hosting plan (because somehow I still haven't got around to migrating the Wongery to the new hosting plan I purchased back in December) and it wasn't working either. Okay, so the problem was with my webhost. This was not the first time my site had gone down at a particularly inconvenient time, or the worst; it had also briefly gone down just after the hard launch in January, which I initially thought was because of the site migration, but no, it turned out they hadn't migrated the site at all and it had apparently gone down for unrelated reasons. In that case, however, when I contacted my webhost they were able to restore the service relatively quickly (despite having completely bungled the site migration). So, conveniently not being at work at the moment (I did work yesterday, but unusually I had a several-hour break in the middle of the workday, and my worksite that day was close enough to my apartment I figured I'd just go home for a few hours, and it was during that break that the site went down), I figured I'd contact my webhost's support.
Before continuing with the account of my trying to get the site back up, however, I'm going to jump back to an earlier event, the relationship of which to the situation described in the previous paragraph I will not immediately elucidate, though I don't think it's difficult to infer.
The following is an excerpt from the body of an email I received about a week ago:
We appreciate the opportunity to serve your online needs. We want to let you know that the registration period for the service(s) listed below has expired. Immediate action is required if you intend to continue using the service(s).
...
Product Description Qty SiteLock Essentials 1 The service is currently in an inactive state. However, the service may still be available to renew and therefore reactivate within your account.
Okay, so apparently my subscription to something called SiteLock Essentials had expired. I had no idea what SiteLock Essentials was. I did not recall ever subscribing to anything called SiteLock Essentials. So I didn't see any urgent need to renew it. I did do a quick websearch to try to figure out what SiteLock Essentials was, and as far as I could tell it had something to do with protecting sites from malware. Well, I'd never had a problem with malware on any of my sites; it didn't seem like this was something I needed to deal with right away, if at all. So I ignored the email and went on with my day.
Now, to resume the previous narrative...
So I found the link to the chat support
I was left waiting for long enough for a representative to join the chat that I figured I may as well call the webhost's phone support as well. This turned out to be trickier that it should have been, because the chat window covered the part of the page where the number for their phone support was listed, and there didn't seem to be any way to move it. I eventually found that number with a websearch, however, and called it. This turned out to be a fortuitous decision, not only because even once a representative finally did join the chat, they were so slow to respond that the phone support had already diagnosed my issue and put its resolution in motion before the chat representative got around to giving me any explanation for the site's downtime, but more importantly because even once the chat representative did give me an explanation, that explanation was wrong. What the chat representative said was that I needed to "optimize the website" because it was exceeding the server resources. This was... very plainly not the case. By far the most resource-intensive software running on the site was MediaWiki, and it seemed unlikely that that was more than the server could handle. Moreover, the other site that wa experiencing similar issues didn't even have that; there were custom scripts on its homepage, but they weren't resource-intensive... and, moreover, it had been almost six months since any of those scripts had been changed, so it made no sense that they would suddenly be causing issues now. This was patently not an optimization issue, but the representative repeatedly insisted that this was the case and recommended I contact a developer, and I probably would have been really confused had the phone representative not already by this point told me what the issue really was.
(Okay, technically I had made a few recent changes to the Wongery website, and specifically to the code for the blog: adding brief descriptions to the posts on the main blog page, adding a link below each blog post to the corresponding forum thread, and tweaking the CSS to fix how certain elements showed up. But first of all, certainly none of that took up a lot of resources, and secondly, I had made those changes the previous day, so they shouldn't suddenly be causing issues this much later. And, again, the other site that was also experiencing issues hadn't had any changes in almost six months. So recent changes to the site code were definitely not to blame.)
It is tempting to conclude from this that in the future I should just contact the phone support, and not try to use the chat. However, this is far too small a sample set to draw any firm conclusions from. It could be that I just happened to get an especially incompetent chat representative this time around, and that usually the chat support is more helpful. On the other hand, I think the completely foozled site migration was arranged mostly through chat, so... hmm.
So what was the real issue, then? Well, as surely nothing in the preceding paragraphs could possibly have given you cause to guess, the site had apparently been infected with malware. And to remove the malware... I would have to be transferred to a representative of SiteLock.
Oops.
I did remember, at this point, that email from a week ago, and realized that I did not currently have a SiteLock subscription. Presumably I would have to have a subscription before they would help me. And indeed, the SiteLock representative transferred to me to yet another representative who would help me renew the service, although this proved unnecessary, as I managed to renew it online. (It cost almost a hundred dollars for the year, but eh, I guess if it's necessary to get my site back up, and keep it up...) Anyway, I was transferred back to the first SiteLock representative, who, after confirming that I now had an active subscription, said they would open a case and that the issue should be resolved within six hours.
(It was only after all of this had already taken place that the chat representative finally responded with their erroneous claim that I needed to optimize the site. By this time I really had to leave to get back to work anyway, though.)
The issue was not in fact resolved in six hours. I didn't even get an email about the case being opened until more than seven hours later. However, within a little over an hour from that email, I got another email saying that the malware clean was completed, so... okay, it was a little more than six hours, but close enough, I guess. The email said that I did have other infected websites not covered by the SiteLock services that it recommended also addressing, but, uh... the SiteLock services should cover the entire hosting plan, so these other domains ought to be included in it as well. I guess I'll be giving SiteLock tech support a call.
(I'm not sure if there's something going around; when I initially went to check my email to see if I'd received a message about the case yet, Gmail appeared to be down. So to check whether Gmail was actually down or whether the problem was just on my end, I went to another site I sometimes used for that purpose, isitdownrightnow.com—but, ironically, isitdownrightnow.com was down! Another site, Downdetector, however, did say that "User reports indicate possible problems at Gmail", so... have there just been a lot of attacks on websites today? Or is it just a coincidence? Okay, probably just a coincidence.)
So, uh, yeah. Apparently SiteLock Essentials really was a bit more essential than I realized...
Anyway, that other blog post that I'd been about to put up yesterday when the site went down will be going up later today.