1

The SLHost Saga

Fahim and I have had a VPS – virtual private server – with SLHost.com for a while. A year and a half, maybe two years? In the beginning, the service was good.

The last half year, however, it’s gone downhill. Fast.

We experienced some problems with the server about a half year ago, so SLHost moved us to a new server. That should fix things, right? Except that it didn’t. We continued to experience some of the same problems. And more.

Our email would be shut down. Fahim would have to turn it back on. Cron jobs shut down. This script or that shut down. Server unresponsible. Fahim would have to either restart whatever services had been shut down or restart the server. Repeatedly.

Not once when any of our services were shut down – by SLHost.com staff  (ETA: I assume, but have no hard evidence of that) – were we informed that they were doing so. Every single time we discovered a service shut down or something not working, it was because Fahim or I discovered it, not because we were told.

The last month it got worse. Services being shut down more frequently with no explanations. Fahim would enter a ticket to find out what was going on. He didn’t get any reasonable answers. (Oops – I was wrong!)

As I type this, by the way, our server is down again. Again. It spent the last 36 hours down – I’ll get to that a bit later – and has only been up for a couple of hours, and again, it’s down.

Email, for both of us, has been spotty with SLHost. I’ve had trouble downloading the email to my notebook for the last month. I can access it through IMAP, but not POP3. But I can’t even move it to my notebook through IMAP, although I can, most of the time, read it and delete those emails that I no longer need. But the emails I DO need I cannot move to my notebook. (ETA: The IMAP issues could be caused by the SLT Internet problems. I can read the email, but cannot copy to a folder on my local machine.)

I have not been able to upgrade any of my WordPress blogs automatically for at least a half year, some for as long as a year. I get out of memory error messages, despite the amount of memory WordPress trying to use being dramatically lower than the amount of memory we supposedly have allocated to our VPS. Sometimes, I get out of memory error messages trying to simply load a blog post or a comments page. Sometimes, it just doesn’t load with no apparent reason.

[blackbirdpie url=”http://twitter.com/LMAshton/status/13119659364581376″]

[blackbirdpie url=”http://twitter.com/LMAshton/status/13119936025075712″]

[blackbirdpie url=”http://twitter.com/LMAshton/status/13120912761028608″]

The last month, I’ve been running into disk space usage problems where my remaining disk space is suddenly used up with no warning to me. Upon investigation, I discover over 400MB of core dump files taking up all of what used to be my remaining space. Core dump files appear when there’s a really nasty server problem. And those core dump files means that email no longer works at all. So I delete the core dump files just to get my domain and email working again. But then, the core dump files – another 400 MB of them – appear again in less than 24 hours. Delete. Reappear. Delete. Reappear. Delete. Reappear. Delete. (ETA: Core dumps appear to be created when an app on the server crashes.)

All of which means the server is suffering some kind of catastrophic failure. At least, that’s how I read it, since my domains were down more than they were up.

Fahim  opened a ticket. SLHost responded that they would move us to a new server.

[blackbirdpie url=”http://twitter.com/LMAshton/status/13456234854227968″]

[blackbirdpie url=”http://twitter.com/LMAshton/status/13456539100643328″]

Except the transfer didn’t work. At least, that’s what we think happened, although they couldn’t be bothered to tell us. Fahim opened another ticket when our servers had been inaccessible for 16 hours – they said the server transfer would take 7 to 8 hours. Their response was only “we’ll look into it.” It’s not like they ever answer actual questions.

[blackbirdpie url=”http://twitter.com/LMAshton/status/13593290905493504″]

[blackbirdpie url=”http://twitter.com/LMAshton/status/13757952049545216″]

So finally our server is back up again after 36 hours of being completely inaccessible. No email, no websites, no back end access. Completely down. And it turns out they couldn’t transfer us to a new server because the old server was running old software which wasn’t compatible with the new server. So our sites were down for 36 hours… WHY?

Of course they don’t answer that question, either. Fahim opened another ticket and requested that they leave our server alone. No more attempts to transfer. We need access to our server so we can remove the data and set up an account elsewhere that will actually function. Meanwhile, Fahim also requested that this be escalated to a supervisor. Of course, however many hours later it’s been, we haven’t heard back from anyone.

Meanwhile, AFTER the domains are “restored” to the old server, after 36 hours of being inaccessible, they again go down. Email still doesn’t work. (ETA: Fahim says that email hasn’t actually gone down, it only looks that way to me.) Up, down, up, down, up, down. Except that even when it’s up, it’s not really working properly.

The whole reason they were attempting to move us to another server was because our existing server sucked. Rocks. Because our experience was terrible. So of course we’re completely down for 36 hours. Sure, so they gave us an extra month free, but why would we want to use it when their “service” is like this?

If you’re considering going with SLHost.com, my question would be… Why?

UPDATE:

We got our Rackspace account and Fahim has moved everything over. Meanwhile, I’ve been waiting to do the WordPress 3.0.3 update on my blogs because with SLHost, I’ve had to do it manually. I’d always get an out of memory error. On the new server at Rackspace, Fahim allocated 64 MB of RAM to PHP instead of the 32 MB of RAM for PHP he had allocated at SLHost. Okay, double the RAM, good. So I figured I’d try doing the automatic updrade on the blogs. The automatic upgrades that hadn’t worked in how long?

And, because I pretty much like pushing things to see what the limits are, I did four blogs at the same time. Because I wanted to see what would happen. You know what happened? The upgrades worked. All of them. With NO out of memory error messages. With NO error messages of any kind. And it was fast. Four. At a time. When I couldn’t do one at a time before.

Yeah. Moving to Rackspace? Good. Decision. Ditching SLHost? Another good decision. Yup. 🙂

UPDATE A SECOND PART, THE FOLLOWING MORNING:

You know how I mentioned I couldn’t move my email from the IMAP server to the local folders on my notebook? And you know how I said it could be because of our crappy Internet connection? Well, guess what?

This morning, the day after the move to the new server at Rackspace, I tried moving the email to my local folders on my notebook. And it worked. Flawlessly. On the same lousy Internet speeds. Yeah. Rackspace for the win. 🙂

1 thought on “The SLHost Saga”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.