Yesterday (March 10th), the server which my old blog was hosted on got melted.
Non, les données contenues dans les serveurs de SBG2 ne seront vraisemblablement pas récupérables.— xgarreau 🤝 (@xgarreau) March 10, 2021
Mais si vous utilisez le Backup FTP #OVH, vos sauvegardes sont dans un autre datacenter. (Les photos sont des pompiers) https://t.co/tRonks8qIw pic.twitter.com/EL5BcMjbc7
To clarify, my VPS was being hosted in SBG1 which was right next door to the the building which caught fire (SBG1). As a precaution, OVH disabled the entire Strasbourg datacentre including SBG1, SBG2, SBG3 and SBG4. I don’t actually know if my server got toasted or not, but since there’ll be an investigation etc, I won’t know for a while.
The blog was running on Ghost v1.x at the time of it’s demise, yes, long overdue an upgrade. I’m pretty sure the server was running Ubuntu 14 as well - whoops.
It had around 50 posts dating back to 2014, so losing it would’ve been a shame.
Thankfully, I had made an export on 19th Sept so I had all the content, bar the latest post. Unfortunately, the export from Ghost didn’t contain the images for the articles so they’re still missing. I’ll have to go through all the old posts and find out which ones have missing images.
I was intending to rip everything out anyway, but hey, best laid plans… not.
We wrote a URL shortener a long time ago which was being hosted on my server. The code is on GitHub but the database was in a MongoDB instance which didn’t have any automated backups, so we’ve probably lost all that data.
It gets pretty minimal traffic so I’m not overly concerned about it.
I figured it was time to try something new for my site, so I’ve gone with Hugo deployed to Cloudflare Pages.
It was remarkably simple to setup once I’d converted the Ghost json export into markdown files for Hugo. I ended up using this tool which works nicely: ghostToHugo
Then it was a simple matter of linking Cloudflare Pages to the new GitHub repo and letting it do it’s thing. I also migrated over to Cloudflare nameservers for the entire domain, rather than using the default Namecheap ones. The site should be a fair bit more performant and reliable now, and I also don’t need to worry about renewing LetsEncrypt certificates either. Win-win!
Everything worked out okay in the end, but it was a good reminder that I need to do a little more due diligence when it comes to storing data, even though it’s only personal stuff.
#hugops for the OVH team, we’re all glad no one was injured.
Many thanks for all the empathy messages you sent us today ! I want to thank the teams who have been working all night/day.— Octave Klaba (@olesovhcom) March 10, 2021
It’s been the worst day for the last 22y and there is no word strong enough to say how sorry I feel today.
We keep working hard to restart SBG1/3/4 asap!