| 1 Jul 2025 |
 djacu | Hey Infra Team
In case you haven't seen the recent post on discourse, the Marketing Team is preparing this year's community survey. I am reaching out to teams to see if there are any questions they would like to add to the survey to better serve the work you all do. More details in the post linked below.
https://discourse.nixos.org/t/community-feedback-requested-2025-nix-community-survey-planning/66155 | 03:27:21 |
 Jonas Chevalier | One question, but that's more of a personal one; would people be interested in being able to track their personal usage of the cache? We could introduce an optional API key system that would enable that type of accounting. | 08:44:10 |
 emily | to enable better analytics for decisionmaking or for consuming by yourself? | 09:06:42 |
| emily | for the former, I think there'd be more uptake if effort was put into doing it in a privacy-preserving way like the work done by Go and Divvi Up | 09:06:50 |
| emily | (actually, this might make no sense, since the cache already sees every request and who is making it, and totally-anonymized stats are already published) | 09:35:48 |
| emily | (but I can imagine cases where you might want to combine stats about requester with stats about requests. but privacy-preserving analytics ingestion stuff is probably not relevant here given the architecture of the cache. publication is another matter) | 09:36:22 |
 Emma [it/its] | wouldnt it make more sense to do this on the client side? | 09:36:33 |
|  sinan changed their profile picture. | 12:07:52 |
| 2 Jul 2025 |
 Robert Hensing (roberth) | for organization-wide aggregation would be a useful feature if implemented centrally. Also the data could be more accurate, taking into account where the cost was made: CDN or backend, possibly including regional differences | 12:40:46 |
| Robert Hensing (roberth) | (not on my personal radar, but could be beneficial for the ecosystem) | 12:41:47 |
| emily | most organizations with that kind of need are probably using their own caching proxy if they're using the central cache at all, surely? | 12:42:03 |
| emily | (so already have a central point to collect data) | 12:42:13 |
| Robert Hensing (roberth) | depends. If we want to motivate businesses to contribute to the infra, we should make the info part as easy as possible | 12:43:59 |
| Robert Hensing (roberth) | otoh some vagueness about the actual use might be good. Ideally they'd pay for value, not cost | 12:44:46 |
| emily | I see, I didn't know the idea was to sell paid services | 12:45:42 |
| Robert Hensing (roberth) | not really. The cache is meant for everyone. As far as I'm concerned, contributions remain optional | 12:46:44 |
| Robert Hensing (roberth) | (and I don't think anyone disagrees with that) | 12:47:04 |
| Jonas Chevalier | emily: Personally I'd be interested in knowing how much bandwidth my personal infra, and Numtide are using. I would be interesting to me. I'm wondering if other people would like to have this facility too. | 13:57:20 |
| Jonas Chevalier | It could also be useful to know which org to approach and ask if they would be willing to install loca caches / reverse proxies to reduce the load. | 13:59:29 |
 Vladimír Čunát | I'd expect the last bit to be covered by Fastly, but who knows. | 14:15:12 |
| Jonas Chevalier | We can already identify orgs with public IP ranges / AS as those appror in the fastly logs. | 14:31:00 |
 ners | In reply to @emilazy:matrix.org
most organizations with that kind of need are probably using their own caching proxy if they're using the central cache at all, surely? AFAIK, the caches I know, such as Cachix and Attic, will not cache locally things that are available in the global cache. | 16:27:14 |
| Jonas Chevalier | I'm expecting GitHub Action runners to use most of the traffic | 16:37:14 |
 Tom | In reply to @ners:nixos.dev
AFAIK, the caches I know, such as Cachix and Attic, will not cache locally things that are available in the global cache. For attic: you can disable the upload filter and then it's only a question of having the priorities configured correctly to primarily use your cache. | 16:48:50 |
| 3 Jul 2025 |
 K900 | https://hydra.nixos.org/build/301602245 might be stuck | 07:36:40 |
| K900 | Yeah definitely stuck | 07:57:41 |
| K900 | I'll just start a new eval instead | 07:57:58 |
| 4 Jul 2025 |
 dgrig | What's the proper procedure for me to request I get write access to Grafana so I can upload a dashboard for the tracker.security.nixos.org service? The dashboard in question is https://github.com/Nix-Security-WG/nix-security-tracker/blob/main/contrib/grafana-dashboard.json if it's easier for someone else to import it. | 11:35:26 |
| dgrig | * What's the proper procedure for me to request I get write access to Grafana so I can upload a dashboard for the tracker.security.nixos.org service? GitHub issue?
The dashboard in question is https://github.com/Nix-Security-WG/nix-security-tracker/blob/main/contrib/grafana-dashboard.json if it's easier for someone else to import it. | 11:35:43 |
| K900 | https://grafana.nixos.org/d/beo2uotj65lvkb/nix-security-tracker?orgId=1&from=now-6h&to=now&timezone=browser&var-Instance | 11:36:40 |
