I’d planned to do a bit of blog admin this morning including upgrading all my WordPress blogs to the latest version 2.3.3, released earlier this week.
Plans scuppered, though, by a long service outage at DreamHost, the company I use for hosting.
Looks like either a continuation of yesterday’s outage (which is what I thought) or, more likely, related to moving a server cluster.
Whatever the reason, it meant that I couldn’t get into any blog, either through the usual login or via FTP. And of course no blogs were visible on the net at all so apologies for that if you came visiting today.
While such outages do upset a lot of people (take a look at some of the comments on the DreamHost status blog post), I tend to take a philosophical view mainly because outages like this are few and far between (at least, in my experience), and DreamHost has a good record of a) communicating at times like this via their status blog and b) fixing things relatively quickly.
The outages last night and today seem to be the worst since a major outage in February 2007 that saw no service for 12 hours or more.
Today’s has been six hours according to the up/downtime reports I get by email from Hyperspin, a useful service that tracks my sites’ availability on the net.
Overall, though, DreamHost does have a pretty good uptime record.
The latest monthly report from Hyperspin for January shows that this site had been available for not far short of 100% of the time.
This level of uptime has been consistent in almost every month during the past year.
I don’t know about you, but I reckon any hosting service that consistently gives you over 99% uptime availability is very good indeed.
So I’m not throwing in the towel on DreamHost in spite of this latest issue. And no doubt the Hyperspin status report for February will show a number much less than 99%.
Still, I must admit that my confidence in DreamHost has been dented a bit. The most alarming aspect of today’s outage was being unable to access anything on the server via FTP.
So, for instance, I couldn’t get at any content in order to copy it locally or to somewhere else.
Earlier, I had a short video conversation via seesmic with Phil Campbell who had some sensible suggestions involving server mirroring. That’s a relatively complex step to take, more suited to a large enterprise or for something that truly can be described as mission critical.
No, I’ll stay with DreamHost.
But ask me if those plans are the same when the next outage happens.
I was with Dreamhost but tired of telling them my server was down.
I am now with Bluehost and have been for 15 months. Enough said.
The recent Dreamhost jokey style letter ‘apologising’ for the previous outage absolutely appalled me.
That DreamHost blog post on the billing issue did offend a lot of people, Andy. DH did subsequently apologize.
For every 5 people who think DH’s service is great (which does include me), I can find 5 others who think it’s awful. (Heh! Same goes with Virgin Media, but that’s another story.)
I’ll stay. But as I said in my post, that’s no longer a permanent state of mind.
Neville, I had the same problem and was intending to do some back end work yesterday. It got thinking about 2 things:
1. Set up my site locally so that I could work on technical stuff and general housekeeping, and even testing new ideas/designs
2. I’ve asked DH about the prospect of setting up a page explaining to visitors that the site is undergoing maintenance and thanking them for their patience (rather than an ominous server error message!)
A nice touch would be if they could brand that notice page for our sites.
At least if they did that I would feel a little bit better about the whole thing and so too I feel, would my visitors.