| [04:01:09] | * spitz234[m] has joined #farmos |
| [10:21:17] | * firgoodway[m] has joined #farmos |
| [10:22:39] | <symbioquine[m]> | Good point/question! |
| [10:27:33] | <symbioquine[m]> | Probably not representative, but it looks like my production system is using about 240 MB of ram right now. 150 MB for the web container + 90 MB for the database container. |
| [10:29:07] | <firgoodway[m]> | that helps a lot, thanks! |
| [10:29:31] | <symbioquine[m]> | I'm trying to generate some more load now to see what that does... |
| [10:32:46] | <FarmerEd[m]> | I run it on a Raspberry Pi with 1GB without much issue. Only ever has 1 or 2 users + some API use. |
| [10:34:29] | <FarmerEd[m]> | Running Node Red on the same device for interacting with API + a few other lightweight services. |
| [10:38:04] | <symbioquine[m]> | Hitting the root api schema (should be a highly cacheable response - but I don't know how much it's actually being cached) with 10 qps bumps the memory usage up to around 285 MB |
| [10:39:49] | <symbioquine[m]> | fortio load -qps 10 -t 30m "https://example.farm/api/schema" |
| [11:02:37] | <symbioquine[m]> | With 10 qps it used about ~8s of cpu time. |
| [11:02:37] | <symbioquine[m]> | For 10 idle seconds, it looks like the web container is using about ~700 usec of cpu time. |
| [11:03:35] | <symbioquine[m]> | Obviously probably won't translate between machines super well, but that was run with a "Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz" |
| [11:04:18] | <symbioquine[m]> | * Obviously probably won't translate between machines super well, but that's with a "Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz" |
| [11:05:20] | <symbioquine[m]> | To get those numbers, I found my container's long-form id with docker ps --no-trunc --format "{{.Names}}\t{{.ID}}" |
| [11:06:23] | <symbioquine[m]> | Then did;... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/bab0cbe6b9...) |
| [11:07:24] | <symbioquine[m]> | I just ran it twice manually for each case - so it's a pretty low sample size. YMMV 😅 |
| [11:25:44] | <symbioquine[m]> | <firgoodway[m]> " * I saw the "server requirement..." <- > <@firgoodway:matrix.org> * I saw the "server requirements" section, but didn't find any hard requirements there |
| [11:25:44] | <symbioquine[m]> | > https://farmos.org/hosting/install/#server-requirements |
| [11:25:44] | <symbioquine[m]> | We should probably include a recommendation on that page of at least 512 MB ram and a relatively modern dual core cpu. |
| [11:36:55] | <firgoodway[m]> | <symbioquine[m]> "I just ran it twice manually for..." <- interesting metric there |
| [11:37:31] | <firgoodway[m]> | thanks for digging into that! That was much more of an involved answer than I anticipated |
| [11:37:51] | <symbioquine[m]> | np |
| [11:38:20] | <symbioquine[m]> | firgoodway[m]: Yeah, there's probably a better way to measure it 🧐 |
| [11:39:01] | <symbioquine[m]> | but at least it seemed like an improvement over "% cpu" |
| [12:15:14] | <mstenta[m]> | > We should probably include a recommendation on that page of at least 512 MB ram and a relatively modern dual core cpu. |
| [12:15:14] | <mstenta[m]> | makes sense to me! |
| [12:39:30] | <mstenta[m]> | Thinking about tagging 2.1.2 soon... |
| [12:40:05] | <mstenta[m]> | w/ new Drupal core version, couple of contrib updates, and the bugfixes etc that we have in 2.x currently |
| [12:40:08] | <mstenta[m]> | Nothing too bug |
| [12:40:12] | <mstenta[m]> | big |
| [12:42:05] | <mstenta[m]> | paul121: got any PRs in the works I should wait for? |
| [12:42:22] | <mstenta[m]> | I'll go ahead and merge https://www.drupal.org/project/farm/issues/3312126 |
| [12:42:45] | <mstenta[m]> | And run tests against new Drupal 9.5.10 release in a moment here... |
| [13:19:37] | <paul121[m]> | <mstenta[m]> "paul121: got any PRs in the..." <- nope! |
| [15:43:26] | <mstenta[m]> | farmOS 2.1.2 has been released! 🎉 https://github.com/farmOS/farmOS/releases/2.1.2 |