kosmos/chef
8
3
Fork 0
Code Issues 39 Pull Requests 1 Projects 2 Activity
2,156 Commits 8 Branches 0 Tags
663c4da0a06d6c8ce2ca8b6328d04d4580d9ba87
Commit Graph

2156 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
greg 663c4da0a0 Update the prometheus-1 node after running Chef 
It now contains the attributes from the ohai plugin
 2026-07-11 17:35:04 +02:00
greg 010e2d0d64 Add zerotier ohai plugin to the base role 2026-07-11 17:34:39 +02:00
greg 065eac06b1 Add a recipe to install the zerotier ohai plugin 
The attributes are saved as "automatic" attributes and look like this:

  "zerotier": {
    "version": "1.16.2",
    "node_id": "NODEID",
    "networks": {
      "NETWORKID": {
        "network_name": "networkname",
        "mac": "ff:ff:ff:ff:ff:ff",
        "status": "OK",
        "type": "PRIVATE",
        "interface": "abcdefg",
        "addresses": [
          "10.1.1.255/24"
        ]
      }
    }
  }

Refs #263
 2026-07-11 17:29:31 +02:00
greg 1759af98b8 Simplify and improve chef attribute handling 
Set the attributes in a Ruby file under /etc/chef/client.d/ instead of
editing /etc/chef/client.rb
Now uses the new whitelist attribute:
https://docs.chef.io/client/19/cookbooks/attributes/attribute_persistence/

Local chef config also uses the same file. That means there is no longer
a need to use the sanitize script to remove override and default
attributes from nodes.
 2026-07-11 17:28:03 +02:00
raucao 736d4effa4 Upgrade RS discourse, increase service timeouts 2026-07-10 11:30:11 +02:00
raucao 1ebe0e4074 Merge pull request 'Update akkounts cookbook, deploy new akkounts VM' (#645) from chore/akkounts into master 
Reviewed-on: #645
 2026-07-10 09:29:44 +00:00
raucao 9250b2c726 Deploy new akkounts node, remove old one 2026-07-10 11:27:01 +02:00
raucao 6807b8f5bc Update akkounts cookbook, deploy new akkounts VM 
* Fix: add missing directory resource
* New nodejs cookbook usage
* Remove Redis entirely (not using local Redis anymore)
 2026-07-10 11:24:42 +02:00
raucao 0338e56d4c Merge pull request 'Upgrade nodejs cookbook from 7.3 to 11.0' (#644) from chore/upgrade_nodejs_cookbook into master 
Reviewed-on: #644
 2026-07-10 09:23:02 +00:00
raucao bd193567ca Upgrade nodejs cookbook from 7.3 to 11.0 2026-07-10 11:19:43 +02:00
raucao 86f7b68726 Merge pull request 'Set up Prometheus and Alertmanager' (#632) from feature/prometheus_cookbook into master 
Reviewed-on: #632
 2026-07-09 13:33:38 +00:00
greg 51af9c76e0 Fix node attribute 2026-07-09 15:30:38 +02:00
greg f15d5cb958 Move the alertmanager port to an attribute 2026-07-09 15:23:46 +02:00
greg ae14a1f353 Move the node_exporter port to an attribute 2026-07-09 15:18:06 +02:00
raucao b700baac37 Configure alerts for disk/filesystem usage, memory, CPU 2026-07-09 14:04:27 +02:00
raucao 6f93820936 Update node configs 2026-07-07 16:58:36 +02:00
raucao 9fca5386f3 Increase padding for NodeDown rule 2026-07-07 16:57:22 +02:00
raucao b6ba5ba679 Update node configs 2026-07-07 15:20:48 +02:00
raucao 4976888358 Remove unnecessary include 2026-07-07 15:20:33 +02:00
raucao e4ad52d413 Update dirsrv node configs 2026-07-07 15:05:25 +02:00
raucao 94a3c41964 Show severity for new alerts, number of firing vs resolved 2026-07-07 14:26:21 +02:00
raucao 704d0bca0e Document upgrading Chef client 2026-07-06 18:50:07 +02:00
raucao 02dc95b7de Update node configs 2026-07-06 18:47:51 +02:00
raucao 2a14837f64 WIP Add prometheus rules, improve alerts 2026-07-06 18:01:22 +02:00
raucao 1ae5210813 Send critical alerts via email (in addition to XMPP) 2026-07-06 14:52:17 +02:00
raucao 20f67d7261 Change prometheus scrape/eval intervals 2026-07-06 14:52:10 +02:00
raucao 4ce77e1a6e Fix postfix relay sending (on newer Ubuntu versions) 
Requires relay restrictions to be configured explicitly
 2026-07-06 14:50:49 +02:00
raucao 21513e6d10 Critical icon only for critical severity 2026-07-06 13:53:52 +02:00
raucao 33987402e3 Fix port attribute 2026-07-06 13:16:10 +02:00
raucao 81aae3d1bb Add icons to visualize alert severity 2026-07-06 12:38:07 +02:00
raucao 65c1933a99 WIP Add default alertmanager route and receiver 2026-07-06 12:38:07 +02:00
greg 5e7a0c07b0 Move web port to an attribute 2026-07-05 18:16:18 +02:00
greg 8601e33764 Fix year in license file 2026-07-05 18:03:12 +02:00
raucao c736468259 Merge branch 'master' into feature/prometheus_cookbook 2026-07-05 17:39:08 +02:00
raucao 6370069e9d Add new vm_host attribute/value to all VM nodes 2026-07-05 17:32:43 +02:00
raucao 0e9bcca72d Add vm_host label to prometheus targets 2026-07-05 17:32:14 +02:00
raucao c762dfe8c1 Merge pull request 'Sanitize node configs via script, and do it automatically in a Git hook' (#635) from dev/sanitize_node_configs into feature/prometheus_cookbook 
Reviewed-on: #635
 2026-07-05 15:26:30 +00:00
raucao 53cf95c4e7 Remove decommissioned nodes 2026-07-05 17:22:48 +02:00
raucao 959d506cfb Remove newline from all node files 
Was actually added by manually editing the files, but Chef writes them
without newline, so our sanitization script also does it that way
 2026-07-05 17:22:39 +02:00
raucao 8c71261fc9 Add pre-commit hook for sanitizing node configs 2026-07-05 17:22:34 +02:00
raucao 6c3da8a711 Add util script for sanitizing nodes 
Deletes the "override" and "default" properties, which may contain
sensitive data
 2026-07-05 17:22:26 +02:00
raucao 53de8ea119 Merge branch 'master' into feature/prometheus_cookbook 2026-07-05 17:20:48 +02:00
raucao 639800a838 Remove decommissioned node 2026-07-05 17:19:16 +02:00
raucao 1e2ff76797 Merge pull request 'Move allowed node attrs to config file, apply automatically' (#634) from dev/allowed_node_attrs into master 
Reviewed-on: #634
 2026-07-05 15:11:22 +00:00
raucao 301565b910 Move allowed node attrs to config file, apply automatically 
This moves the whitelist for nodes' "normal" attributes to a config
file, which is loaded live during every run and applied in memory in
case the node's local file hasn't been updated before the client run.

Fixes allowed attributes being overwritten/removed in the node info
JSON files.
 2026-07-05 17:07:56 +02:00
raucao 171db0fb7c Add altermanager section to prometheus config 2026-07-05 13:19:18 +02:00
raucao 821b56e6ab Set node name as prometheus instance name 
And DRY up the recipe so it's easy and expressive to add more
metrics/targets
 2026-07-05 12:41:46 +02:00
raucao e29d90d3d0 Use full domain as fqdn for leo 2026-07-05 12:40:50 +02:00
raucao f2ebda4a1a Update node configs 2026-07-04 16:30:15 +02:00
raucao 67f62ebd6c Scrape garage metrics 2026-07-04 16:14:36 +02:00