Downtime, Upheaval, and Migration

Ektoplazm has recently bounced back from several weeks of intermittent downtime. I have had some serious problems with my former hosting provider, OpenSourceHost. Earlier this month they moved my site from one server to another during a routine upgrade. In doing so, something snapped. I am not entirely sure what it was–all I know is that things were running smoothly before the upgrade. Afterwards, it was bombing each and every day.

OSH’s support was no help at all. None of their CSRs had a clue. It took me a few calls just to reach someone who knew that there was an upgrade going on at the time. Prior to that I was hearing vague prognistications about DNS propagation and witless platitudes about clearing the browser cache. As it turns out, the .htaccess file in the root did not survive the transfer from one server to another and had to be regenerated. No big deal, but no one told me what was going on–it was a complete information vacuum. Oh, I was livid. And even as that first problem was solved, new issues sprung up to demand my attention. The site became impossibly several times a day. The CSRs eventually settled on a dubious explanation involving backup processes and assured me that the situation was temporary.

It was not. It went on and on, every day for another week. I opened tickets, made calls, provided information, issued complaints, and nothing changed. I began to look into migrating Ektoplazm to another host as a last resort. I was not keen on the idea however–I knew it was going to take a while and unexpected problems would arise. I’ve migrated once before and it was a huge headache. I preferred to leave things as they were and have the damn site fixed. It had been working fine before the “upgrade” after all!

In the middle of all these problems I was able to break through to someone competent; a team leader. Unlike the CSR grunts, he seemed to know what was going on. He applied a refund to my account and assured me that everything would be back on track in a few short days. Of course, nothing changed. And the next time I submitted a ticket they suspended my account entirely. Apparently I was the source of the problems all along. Bullshit! Even if it was, did it really have to take them two weeks to figure that out? A more likely scenario is that they decided it was too much trouble to actually fix my problems. After all, I was on shared hosting–and only paying about $20 a month. I suppose there is a point at which providing proper service is less economical than simply offloading a customer. Evidently my business was not important to them.

This is not how it has always been. Years ago, I would regularly receive email support from the CEO himself. You see, OSH was a mom-and-pop kind of operation, small-scale and specialized in open source content management systems. Over time they have grown and their service has suffered. In the words of a friend, they have outsourced their core competency. I was running open source software on an open source host and the technical staff were absolutely clueless as to what the source of the problem was. What a joke!

So there I was–locked out of my email with no functioning web site to speak of. I called in immediately, burning with rage but willing to compromise. After all, they might have been right! WordPress is a beast. Sites such as mine rest on the threshold of needing a dedicated server solution. Still, everything was fine before the upgrade. And so I asked to be connected to the team leader–who was also responsible for suspending my account in the first place. I was told he was out to lunch and could not be reached. I hung up and left a ticket requesting more information. Not too long thereafter I called in again and was told the team leader–the only person with the authority to lift the suspension–had left for the day.

Game over. I exploded with righteous anger. I had put up with weeks of bullshit service only to have my account nuked. The man responsible avoided me, failed to answer my support ticket, and then left for the day. So I bamboozled the CSR into opening up the account again. Hell, at that point I just wanted to get my data and leave. The very moment I found out about the suspension I had altered the name servers associated with the Ektoplazm domain. The migration was already in motion. All I needed was the data.

I spent the next week moving everything over and working out the unavoidable bugs. The domain propagated quite quickly and within 48 hours normal service had resumed. The new host is a little slower but I’ve done some work to optimize the site and hopefully the difference isn’t noticeable. This whole ordeal has been an immense amount of work. I wish I had been able to invest all that time into something else–it is the end of summer, after all, and I’ve spent almost all of it indoors attached to my computer–but I had no other choice. The spice must flow.

Since completing the migration I have monitored CPU and memory usage on my old host. Unsurprisingly, it shows no real change from when I was still with them. What a bunch of jerks.

Anyhow, these things happen. Better now than when school is on, I suppose. At least it is over–for now. Ektoplazm is beginning to outgrow shared hosting. Soon I will need to look into getting an affordable dedicated server. And migrating again. On my own terms, this time.

So now you know what has been going on. I thank you for your patience and continued support.