Update 2022-04-01-TILvids-Update-April-2022.md

This commit is contained in:
TILvids 2022-04-02 11:45:54 -04:00 committed by GitHub
parent 5121815531
commit 9ecbbfdf1b
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23

View file

@ -3,6 +3,8 @@ layout: post
title: TILvids Update - April 2022 title: TILvids Update - April 2022
--- ---
![image](https://user-images.githubusercontent.com/69435791/161336613-36445ba5-43bb-4d34-9311-725c94d4e6a0.png)
Welcome to April 2022, TILvids community! Lets take a look at what was happening at [TILvids](https://tilvids.com) over the last month: Welcome to April 2022, TILvids community! Lets take a look at what was happening at [TILvids](https://tilvids.com) over the last month:
- So, a funny thing happened on the way to the blog update. I have a calendar reminder set to write up a blog post at the end of the month, but my reminder series went out and I forgot to post last month! I said to myself, "Oh well, it wasn't a big month, I'll just squeeze March and April together, since there's probably not a lot to report..." Famous last words, as it turns out. I noticed toward the last week or so of March that the site was loading really slowly, video thumbnails weren't showing up on social media, etc. I reached out to the Yunohost community (we use their technology to admin the site) and they said that the 4.1.0 release might have some issues, and that updating to 4.1.1 should hopefully clear it up. I usually lag our instance behind a few weeks after updates, just to let the dust settle a bit, but it was bad enough that I opted to roll the dice. Oops. Upon initializing the update procedure, there were a number of errors that occurred. The system ended up bailing on the install, but it left the server in an inoperable state. I had to spend the next few days talking between the PeerTube and Yunohost communities to even track down what was happening. In the end, I had to manually dig into the PeerTube service and delete a plugin to support Chromecast playback that appeared to be no longer compatible with PeerTube, along with deleting over 150 database records that somehow became duplicated. All in all, it took about three days to get to the bottom of what was happening, but happy to report that the server did recover after that, update 4.1.1 was rolled out, and it appears the original problem I was trying to solve did indeed improve after the update. Lots of lessons learned and things to think about coming out the other side of this, but ultimately, just happy everything is back to a working state again. Thanks for hanging with me while I tried to sort it out! - So, a funny thing happened on the way to the blog update. I have a calendar reminder set to write up a blog post at the end of the month, but my reminder series went out and I forgot to post last month! I said to myself, "Oh well, it wasn't a big month, I'll just squeeze March and April together, since there's probably not a lot to report..." Famous last words, as it turns out. I noticed toward the last week or so of March that the site was loading really slowly, video thumbnails weren't showing up on social media, etc. I reached out to the Yunohost community (we use their technology to admin the site) and they said that the 4.1.0 release might have some issues, and that updating to 4.1.1 should hopefully clear it up. I usually lag our instance behind a few weeks after updates, just to let the dust settle a bit, but it was bad enough that I opted to roll the dice. Oops. Upon initializing the update procedure, there were a number of errors that occurred. The system ended up bailing on the install, but it left the server in an inoperable state. I had to spend the next few days talking between the PeerTube and Yunohost communities to even track down what was happening. In the end, I had to manually dig into the PeerTube service and delete a plugin to support Chromecast playback that appeared to be no longer compatible with PeerTube, along with deleting over 150 database records that somehow became duplicated. All in all, it took about three days to get to the bottom of what was happening, but happy to report that the server did recover after that, update 4.1.1 was rolled out, and it appears the original problem I was trying to solve did indeed improve after the update. Lots of lessons learned and things to think about coming out the other side of this, but ultimately, just happy everything is back to a working state again. Thanks for hanging with me while I tried to sort it out!
@ -20,7 +22,3 @@ Here are a few thoughts for May:
- Server migration. While trying to fix the server issue, I had a good chat with some folks on our chat servers, and they tipped me off to some different hosting options that not only have better resources, but also might even reduce costs a bit (which is something that needs to be done to ensure long-term viability of the site). I'm going to investigate that a bit. I don't want to make any changes just yet, coming off of all the excitement from last month, but it's on the roadmap, so expect some changes there at some point. - Server migration. While trying to fix the server issue, I had a good chat with some folks on our chat servers, and they tipped me off to some different hosting options that not only have better resources, but also might even reduce costs a bit (which is something that needs to be done to ensure long-term viability of the site). I'm going to investigate that a bit. I don't want to make any changes just yet, coming off of all the excitement from last month, but it's on the roadmap, so expect some changes there at some point.
I think that's it for now! Thanks so much to everyone for continuing to watch, comment, share, and support the [TILvids](https://tilvids.com) community! ❤️ I think that's it for now! Thanks so much to everyone for continuing to watch, comment, share, and support the [TILvids](https://tilvids.com) community! ❤️
![image](https://user-images.githubusercontent.com/69435791/161336613-36445ba5-43bb-4d34-9311-725c94d4e6a0.png)
(this image taken during a break period from trying to fix the outage, for posterity...)