In case you hadn’t noticed, Twittercism has been down the last few days.
It’s been up some of the time, too, but that was only through sheer perseverance on my part. You see, the problem was caused entirely by my web host, 1&1 Internet UK.
On Tuesday, this site suddenly went dead, with no obvious reason. The blog couldn’t connect to the database that powers it, and when I checked my control panel, I was informed that this was because the database had been closed.
So, I called 1&1. 1&1 are notorious for their bad technical support, but in case you’re not familiar with the process, I’ll lay it out for you here.
There are three tiers of customer support found in most technical organisations. The first level deals with the high-frequency but basic queries that can be solved fairly easily. If level one can’t handle the problem, they bump it up to level two.
Level two represents a smaller but more technically-proficient team than can provide more in-depth support.
Level three is the technical support cream of the crop; there might only be a handful of these guys, but they know everything.
At least, on paper. That’s the system. It’s flawed and ugly but it works for most of the people most of the time, simply because something like 70-80 per cent of all technical problems are solved by the team in level one. Why? Because they are predominately dealing with queries such as, “Does Google have a website?”
It gets awkward when you have a level three problem. It gets really awkward when you know it’s a level three problem, but you still have to start with the people in level one. Here’s what happens:
- You spend 20 minutes on the phone with level one, explaining your problem
- They can’t handle it, so they transfer you to level two
- You spend 20 minutes on the phone with level two, explaining your problem
- They can’t handle it, so they transfer you to level three
- You spend 20 minutes on the phone with level three, explaining your problem
- They (hopefully) fix it
All this on a premium line. But what if they can’t fix it? What if they won’t, because they were the ones who caused the problem?
The awkwardness of this already laboured process is compounded with 1&1 because their system admins (and servers) are based in Germany and their level three support team are in the USA. The former cannot be contacted at all. The latter only after 1pm GMT.
So, on Tuesday, I called technical support. Level one didn’t have a clue. Level two seemed to know even less. Eventually I got pushed through to level three, who also had no idea why my database had been closed because the admin team had left no notes on the system. So they reopened it. Fine, problem solved, I figured.
Wednesday, at around the same time, the database was closed again. More phone calls. This time I actually got level two to give me the direct line for level three (it was still a UK number), and called them directly. Again, no admin notes. Again, they reopened it.
Thursday: the database is once again closed. I wait until after 1pm, and call level three. Now, they’re suddenly reluctant to reopen the database, because even though there are still no admin notes, they feel it’s a bit strange that they’d keep closing my database without a darn good reason.
The level three guy does a scan of my server space and finds a couple of minor things, and suggests there’s a chance Twittercism has been involved in a DoS attack, similar to what Twitter, Facebook et al have been through the last few days. It’s all a bit vague, and even though the ‘fixes’ he recommends are more than a little ambiguous, I say that I will do them. First, however, he needs to re-open the database, which after some reasoning, he does.
I make the changes. I upgrade to the latest version of WordPress, re-optimise the database, delete the files he felt might have been exploited, and away we go.
Sometime this morning, the database was closed again.
I’ll be frank – I haven’t even bothered phoning today. I mean, what’s the point? There will be no admin notes, and there will be no valid reason. At no point during this lousy period did anybody from 1&1 contact me. I was never warned that my database was in jeopardy, or was going to be closed.
Shit. Just. Happened.
So, instead, and with a little bit of research, I moved Twittercism to another database. Thanks to Lester Chan and his fantastic WP-DBManager plugin for making this an absolute breeze. Thanks to me for having the common sense to backup my database on a regular basis. (I’ve actually lost my last two posts, but that’s been because I haven’t actually been able to backup the database.)
Here’s the catch – this new database is still on 1&1’s servers. Currently, it’s running beautifully, but if there actually is some kind of exploit or DoS attack underway then there’s every chance 1&1 might close this new database down, too.
Hence, I’ve been doing some research on moving all of this to a new host, and that’s a project for the near term. I’m really liking eHosting. The dedicated server package they offer is fantastic value for money, and I’ve been really impressed by their customer service. It’s been fantastic. Their online sales team has immediately known the answer to all of my technical questions. Subsequently, each time we’ve communicated I’ve come away happy. (Thanks to James for the recommendation.)
Compare this with the service I’ve received at 1&1, an organisation I’ve been with for years, spending probably close to a thousand pounds. And get this: as a solution to my problem, they actually had the nerve to recommend I upgraded my package, at twice the cost. (Funnily enough, the included link did not work.)
Now, who do you think is going to win my business?