Pandemic Journal, Day 516

We were at the Farmers’ Market this morning for the first time in a month; it’s still local king salmon season, so we had that for lunch today.

I spent the rest of the day catching up – and the rest of this post is basically documentation for Future Me in case I have to prevent SSL certificate expiration or if I want to track down what really made a flight late. Feel free to read along, or to stop here!

While we were traveling, I got an email telling me that the SSL certificate for this blog was going to expire on September 1. This surprised me because I thought I had it set up to automatically renew itself, but I didn’t want to do anything until I got home. Today, though, I took a look at the problem.

First, I had to get rid of the last traces of the temporary domain I’d used while migrating the blog to its new server. That was easy and almost obvious.

The second thing I had to fix was the DNS entry for one of my auxiliary domains – it had an “A” record for the old address of my server, which no longer worked. I changed it to a “CNAME” pointing to the server – that way, it’ll move automatically if I change the address of the server.

The final problem was harder to figure out – the certbot renew command failed with an error message telling me that my account (with Let’s Encrypt) didn’t exist. After a bit of digging, I found that I’d created a new account when I set up the new server but that the old account was still specified in the configuration file in /etc/letsencrypt/renewal – changing the account line in the configuration file fixed that problem, and I was able to renew the certificates.

I won’t really know if I fixed everything until the next renewal attempt in about 60 days, but at least now I’ve documented things for Future Me.

And speaking of documenting for Future Me, I wanted to find out more about why our Boston-LA flight on Friday was so late and whether Delta had actually held the LA-San Jose flight for us.

FlightAware is my usual tool for looking at flight histories – I used its search tool to find the LA-San Jose flight and then used the “Track Inbound Flight” link to go back through all of the flights which that airplane (tail number N289SY) had taken on Friday. Its first flight of the day left Reno more than an hour late, and it never caught up – so it arrived in LA late and Delta didn’t have to hold the departure for us.

I was also curious about what had happened on the Boston-LA leg; FlightAware wasn’t any help. I knew there was a site that showed every event affecting a flight’s timing but I couldn’t remember its name – it took a while to find it again: FlightStats.

I looked at FlightStats’s record of events for that flight (DL346). There were four different airplanes assigned to the flight, starting with N707TW at 3:05am EDT on the 11th. A bit over a day later (4:26am on the 12th), they changed to N722TW. At 11:40am, yet another change, this time to N717TW. And a mere 43 minutes later, another reassignment to N545US (that’s when they downgraded us from Delta One and sent us credit vouchers).

But they weren’t finished making changes yet – at 11:32pm on the 12th, they changed the expected wheels-up time from 11:10am to 4:44pm (oops!) and two minutes later, they changed the airplane to N537US, which eventually brought us to LA.

What happened to all those other planes? I thought you’d never ask. N545US got assigned to the morning BOS-LAX flight on the 13th, which operated on time. N717TW didn’t go anywhere on the 13th (it spent the whole day in LAX – I’m guessing maintenance). N722TW flew LAX-BOS on the afternoon/evening of the 13th. And N707TW had a busy Friday the 13th, flying LAX-BOS and BOS-SEA.

I’m glad I’m not in charge of scheduling airplanes for Delta (or anyone else)!