| 12 Dec 2024 |
 cole-h | In reply to @janne.hess:helsinki-systems.de
To get things started, I clicked a new server (no OS yet so no keys yet) but that gives me the IPs ;) Can someone (not sure who owns the domain RN) add the corresponding records? I thought about going with core01 to make future (but unlikely) scaling possible while not conflicting with the current records to make transition smoother. That would be:
core01.ofborg.org. IN A 138.199.148.47
core01.ofborg.org. IN AAAA 2a01:4f8:c012:cda4::1
Since v6 is actually 2a01:4f8:c012:cda4::/64, I just picked the first one. If anyone has any other preferences, please let me know :D Maybe we can code 0fb0[rg] into it?
Looks like I still have access to the domain (cc @zimbatm:numtide.com -- there's probably an email sitting in the foundation mailbox that someone will need to act on to complete the transfer) so I just added those records I think. 300 TTL just in case. | 21:22:16 |
| 13 Dec 2024 |
 Jonas Chevalier | yep. I'm waiting on Ron to click on the approve link. | 09:46:22 |
| 14 Dec 2024 |
 Janne Heß | Okay, I now have deployed core01 with NixOS. I also created a new repo here: https://github.com/ofborg/ofborg-infrastructure which is very similar to the NixOS infra repo so we can merge it in the future. The new config also pulls the common thingies like SSH keys from the NixOS infra repo to be as similar as possible. Next step would be to provision the new rabbitmq and webhook receiver. Could you, cole-h (if you still own the domain ;) add CNAMEs to core01? gh-webhook and messages (because I don't know if the events. is still is use) should be good enough for me to add nginx and RabbitMQ with TLS certs | 16:24:47 |
| Janne Heß | Also, it'd be nice if I could get in touch with someone from the infra team (hexa?) regarding backups, exporters and loki | 16:25:21 |
 hexa | I can probably give you a storagebox acccount and you can get your exporters queried by adding them in build/pluto/prometheus/exporters | 16:27:45 |
| hexa | no loki yet | 16:27:49 |
| Janne Heß | So we use the pull-based model and not the alloy one? | 16:28:07 |
| hexa | yes, that is where we currently are | 16:28:25 |
| Janne Heß | I will create a PR against the infra repo once I have the rabbitmq exporter set up | 16:28:48 |
| hexa | I don't think that is written in stone or anything, so if there were benefits to migrate to something else we can do it | 16:29:08 |
| hexa | * I don't think that is written in stone or anything, so if there were worthwhile benefits to migrate to something else we can do it | 16:29:16 |
| 16 Dec 2024 |
| cole-h | In reply to @janne.hess:helsinki-systems.de
Okay, I now have deployed core01 with NixOS. I also created a new repo here: https://github.com/ofborg/ofborg-infrastructure which is very similar to the NixOS infra repo so we can merge it in the future. The new config also pulls the common thingies like SSH keys from the NixOS infra repo to be as similar as possible. Next step would be to provision the new rabbitmq and webhook receiver. Could you, cole-h (if you still own the domain ;) add CNAMEs to core01? gh-webhook and messages (because I don't know if the events. is still is use) should be good enough for me to add nginx and RabbitMQ with TLS certs I missed this, sorry! But also it looks like the transfer went through, I no longer have access to it in my Gandi dashboard (so can't add the records) | 17:21:26 |
| Janne Heß | Yaay, but also: oh no 😔. @zimbatm:numtide.com could you? | 17:26:11 |
| 18 Dec 2024 |
| Jonas Chevalier | It looks like the domain is in limbo, I'm not seeing it either | 07:40:33 |
| Jonas Chevalier | Janne Heß: ok, can you give me the new IPs? | 09:10:52 |
| Jonas Chevalier | * Janne Heß: ok (fixed), can you give me the new IPs? | 09:11:16 |
| Janne Heß | In reply to @zimbatm:numtide.com
Janne Heß: ok (fixed), can you give me the new IPs? Ah is the zone empty now? 👀 | 09:14:02 |
| Jonas Chevalier | It kept all of the same entires | 09:15:08 |
| Jonas Chevalier | 
Download image.png | 09:15:28 |
| Janne Heß | In reply to @zimbatm:numtide.com
It kept all of the same entires Ah whew. If the core01 record is still there, you can just add CNAMEs to it so we only have the IPs in a single place | 09:15:58 |
| Jonas Chevalier | Just to be clear: update the @ ALIAS to point to core1.ofborg.org? | 09:17:24 |
| Jonas Chevalier | And all the other CNAME? | 09:17:42 |
| Janne Heß | No, that's the old infra ;) I'm migrating it service by service to the new infra. core ist old, core01 is new | 09:18:46 |
| Janne Heß | That should allow for migrations with minimal downtime, and only for the website (which isn't that critical) and the logs endpoint | 09:19:31 |
| Jonas Chevalier | * Just to be clear: update the @ ALIAS to point to core01.ofborg.org? | 09:20:46 |
| Jonas Chevalier | Ok, tell me exactly which record you want me to update and I'll do it :) | 09:21:03 |
| Janne Heß | Should only be new records for now:
- gh-webook.ofborg.org. IN CNAME core01.ofborg.org.
- messages.ofborg.org. IN CNAME core01.ofborg.org.
I believe cole used 300 TTL
| 09:30:49 |
| Jonas Chevalier | done | 15:14:32 |
| 20 Dec 2024 |
|  🐰 xiaoxiangmoe joined the room. | 13:59:53 |
| 21 Dec 2024 |
|  @stablejoy:matrix.org left the room. | 05:08:24 |
