So last night, a storm knocked out my WordPress setup. It’s running on a Dell OptiPlex 3070 Microdesktop, which is supposed to be protected by a 600KVA UPS. Power was only out for five minutes, but that was enough to take the WordPress server down. Surprisingly, my 2019 Asus laptop, which cost around $400 and has an AMD A12 CPU, wasn’t affected at all—even with its degraded battery. So, clearly, the UPS didn’t do its job.
I also have a Synology drive, which is supposed to create daily backups. Did that work? Well, sort of… it turns out I hadn’t configured it properly to create daily snapshots, so the copy there was also corrupted.
Going through my files, I found a backup script I wrote to compress and back up my volumes. But—guess what? It stopped working. The script was set to keep seven days of daily backups, but it only had one file. Turns out, I had written the bash script incorrectly.
Did I remember to keep a weekly backup? No, I got too busy and overlooked it. What saved the day was an Updraft backup from September 4th, so I only lost about 12 days of posts (and I’ve been posting almost every day!). Plus, I lost all my scheduled posts—what luck.
Despite the frustration, this situation forced me to learn more about handling volumes in Docker. I discovered a method where I could copy the volumes within the image and create multiple volumes. Then, I’d restart the Docker container with different volumes to check if the backup was effective. It’s a neat trick.
This whole process also led me to set up my $50 thin client. After some debugging (aka checking the logs), I found out the MySQL instance wasn’t compatible with the CPU. So, I had to switch to MariaDB.
Anyway, I was just telling Mike how often I break my lab experiments. The upside? Every time I break something, I learn something new, which I can then share with my team.
There was a storm last night and it knocked out
Leave a Reply
You must be logged in to post a comment.