Earlier this evening, at around 7pm London time, there was a brief interruption of service here, lasting around ten to fifteen minutes. This happened during the process of updating the security certificate. I think I have that approximately right. An error was made, and then speedily identified and corrected.
Apologies if you were inconvenienced. The good news is that this is not the sort of problem which is at all likely to happen again.
In general, I hope you will agree, the new blog has been very solid. I expect that to go on being the case.
>The good news is that this is not the sort of problem which is at all likely to happen again.
It’s unlikely to happen again when the security certificate is being updated. Nothing broke due to a bug – I just switched the blog off, and then switched it back on again a few minutes later when I realised what I had done.
As to whether other things might go wrong when doing some other kind of maintenance, well, there is always some risk.
“Nothing broke due to a bug – I just switched the blog off, and then switched it back on again a few minutes later when I realised what I had done.”
Why did Michael switch your blog off?
I did something that I thought would install the new certificate. It didn’t work. I therefore stopped the Apache server, attempted again to install the new certificate, and then restarted the Apache server. It then (for reasons unknown) restarted with a different configuration, meaning that it was showing a static splash page instead of Brian’s blog. I then had to figure out how to change the configuration options back to the correct ones. Once I did that, Brian had his blog back.
It turned out that the certificate was being stored in a non-standard location on this particular combination of Ubuntu/Apache/Wordpress that AWS uses as its standard/recommended WordPress instance. I then wrote myself instructions explaining all this, so hopefully everything will be smooth next time I try to do this.