Important facts about the ongoing #AberdeenCloud situation and how it might affect their customers and our own.
I have already posted the technical details of what went wrong with #AberdeenCloud on Tuesday evening. This is a follow-up with some answers to questions we’ve been asked as events unfold.
Where are you moving us to?
If you’re on Drupal 6 we’re moving you to a spare virtual machine in our private cloud, because we don’t think any current platform supports PHP 5.3, which is the newest version Drupal 6 is happy with. We will not support it for very much longer, but while we can, one of our VMs is the best home for Drupal 6 websites.
If you’re on Drupal 7 we’re moving you to Platform.sh. They provide a similar platform for a similar price point, so it’s a natural fit.
How do I know this isn’t “out of the frying pan, into the fire”?
For Code Enigma, our support and hosting services are profitable and our backups are particularly stringent, so even if anything did happen to us, we would always be able to provide all your server data from AWS S3 buckets, kept in Ireland and well away from our Rackspace facilities. The S3 buckets themselves cost us very little, so we would always be able to keep those running long enough for people to pull their data, though of course, we would never allow our business to go under without giving proper 30 days of notice, as stipulated in our contracts - and we of course hope that would never happen!
As for Platform.sh, they too appear to be well funded and stable. We can’t say much more than that, and that we have carried out our usual background checks on the legitimacy of the business and its registration. We have clearly been bitten badly by the #AberdeenCloud experience, so we do not “endorse” Platform.sh, per se, but it feels solid.
In any case, this experience has taught us to ensure we make it clear to customers using third party platforms, whether managed by us or otherwise, that they really should take their own backups periodically if the data on their website is business critical.
How come you managed to retrieve some of my files?
In a word, luck! Our support developers, old files from when we set you up, residual backups from other events, we managed to hunt out some files. But none of these came from #AberdeenCloud, all files there were totally irretrievable.
How come you managed to rescue our database and code?
Fortunately, it seems the database service is distinct and separate, and still operational. Also fortunately, #AberdeenCloud had provided an alternative means of getting to your databases (via phpMyAdmin) which was still operational, even though your Drupal websites were not. The same goes for the Git version control system, which was still up and running at the time, allowing us to grab your modules and code.
How come my website at #AberdeenCloud is still working?
If that’s the case, it won’t be for long. It is only because there is a cached copy of your site in a reverse-proxy system called Varnish, which is part of the #AberdeenCloud platform, and - like the databases - that Varnish service is still working. But content will not sit in it for ever. At a certain point pages will “expire”, Varnish will try to retrieve a new page from Drupal, it will fail and your site will go down. In most cases Varnish has already failed.
I have my files, can’t you automatically put them back for me?
We can try! We can’t promise anything, but we’re willing to give it a go. Though frankly, in most cases, by the time we’ve written the necessary scripts you probably could’ve uploaded them yourself. Most customers to date are just re-uploading their files manually.
How are you, Code Enigma, involved in this?
We offer (or did, at least) service management services on top of #AberdeenCloud for non-technical clients who do not want to manage the platform for themselves. We did this pretty much at cost, just billed on what #AberdeenCloud billed us. Why? Because some customers do not feel they need the level of SLA and assurance our Rackspace private cloud offers, often simply because that kind of product is more expensive than they are willing to pay for.
That's fine, but in such an instance we are not responsible for maintaining the integrity of the infrastructure, software, backups, etc. - that falls entirely to #AberdeenCloud. The scope of our services for #AberdeenCloud purposes was service management, nothing more. We could never, and did never, claim any control over the infrastructure.
Please don’t read that as a brush-off, we are doing everything we can to help our customers get back up on their feet, but it is what it is. A catastrophic failure of a third party.
Can I sue #AberdeenCloud?
You can try. They’re certainly in breach of their SLA, SLA “Uptime Promise” is on the pricing page, and you could argue they have been grossly negligent in allowing services to lapse without any notice. However, realistically #AberdeenCloud is a Finnish entity, there’s probably no money left in it, there’s probably nobody there to sue. You could go after the executive management, but again, we suspect it would be very difficult to file any legal proceedings against a Finnish individual in Finland because of a corporate dispute. And to what end? It seems to us unlikely there is any recourse to be had.
All that remains to do is reiterate how deeply sorry we are that some Code Enigma customers were affected by this blow-out, and understand we will continue to do our best to support you.