Commit Graph

241 Commits

Author SHA1 Message Date
8b84e0e5b0 Enable Grafana metrics. 2023-08-14 15:03:44 +01:00
d8e24eeb50 Prometheus job config. 2023-08-14 14:58:28 +01:00
97f29680fd Enable nomad metrics. 2023-08-14 14:58:03 +01:00
3fc47a6fa3 Enable consul metrics. 2023-08-14 14:51:48 +01:00
5459aea4fd Enable Consul UI and make it listen on 0.0.0.0. 2023-08-14 14:34:59 +01:00
e65df9e346 Enable default routing for the prometheus metrics. 2023-08-09 15:33:18 +01:00
c788a05c0d Use a nicer default job name for vector metrics service. 2023-08-09 15:29:04 +01:00
09894bdee5 Export host and internal metrics for prometheus, and simplify hostname setting. 2023-08-08 15:52:01 +01:00
eb8c962c4a Only use node as a label, and try harder to set it. 2023-08-08 15:21:14 +01:00
04e8e86ec2 More resources and less aggressive healthchecking. 2023-08-08 15:20:52 +01:00
f36109a0f8 Send journald logs to loki. Not sure if working yet. 2023-08-08 15:08:21 +01:00
8528a8fec7 Send vector internal logs to loki. 2023-08-08 14:46:02 +01:00
7a30c0e9f2 Send glusterfs logs to loki. 2023-08-08 14:38:25 +01:00
ab80d00632 Add dots between job name components. 2023-08-08 14:11:18 +01:00
4cb69f1f83 Replace promtail with vector for loki. 2023-08-01 06:33:50 +01:00
1b2ab9da6e Save traefik logs to loki. 2023-07-21 07:54:00 +01:00
f13b350f73 Add vector service. 2023-07-21 07:16:19 +01:00
496e3c9e44 Switch remaining services from nomad host volumes to docker volumes. 2023-07-21 06:59:32 +01:00
0e7a34a341 Loki service configuration. 2023-07-21 06:31:05 +01:00
3c1fbc7103 Add compute node IPs to /etc/hosts to hopefully make things start up before the network settles. 2023-07-20 07:06:11 +01:00
0dbe33c82f Update flake. 2023-07-20 06:41:29 +01:00
72d651c158 Revert to dhcp assigned IPs. Static ones result in duplicate ping responses (?!). 2023-07-20 06:41:11 +01:00
e8cec993e7 Put glusterfs data on a non-bind mount. 2023-07-18 17:38:35 +01:00
ad697c7108 Update flake. 2023-07-17 21:09:42 +01:00
91f801401e Switch to host volumes. Grafana not working yet. 2023-07-17 21:02:59 +01:00
9a4d00bfd3 Mount the glusterfs. 2023-07-17 12:55:18 +01:00
afca76d949 Keep track of commands run to set things up that aren't in the configs. 2023-07-17 06:53:05 +01:00
f256da7f8c Persist gluster data dir. 2023-07-17 06:42:43 +01:00
cbd56dff8a Temporarily disable binding to tailscale host_network. 2023-07-14 06:50:34 +01:00
3e9c3bc91b Try to force tailscale network to ipv4. Doesn't seem to work. 2023-07-14 06:50:14 +01:00
92374dff89 Rip out seaweedfs and install basic glusterfs instead. 2023-07-14 06:49:55 +01:00
4a6a84190b Switch compute nodes to static IPs to bring up their network faster after a power failure. 2023-07-14 06:49:07 +01:00
a05c4567a4 Update flake. 2023-07-13 05:59:15 +01:00
1e19bd1329 Try harder to wait for the network before starting seaweedfs and nomad. 2023-07-10 13:28:41 +01:00
e98172c572 Only expose API on https port. 2023-07-10 06:38:41 +01:00
73a61bf18a Grafana config. 2023-07-10 06:36:41 +01:00
e3551c1028 Bump max volume limit. 2023-07-10 06:02:02 +01:00
74b0fa9627 Put dashboard behind traefik and enable metrics. 2023-07-10 05:27:21 +01:00
c7b229ff8e Default to letsencrypt for traefik services. 2023-07-09 19:44:58 +01:00
155c584c82 Use keepalived to set up a virtual IP for traefik. 2023-07-09 15:45:58 +01:00
b8eb6dbe77 Switch to unstable package and enable all docker capabilities. 2023-07-09 15:45:46 +01:00
b528f7e3aa Add traefik service. 2023-07-09 14:25:34 +01:00
3a5c96c9b4 Stop using old middleware. 2023-07-09 14:13:18 +01:00
e2359ce283 Move webserver to tailscale net. 2023-07-08 20:05:52 +01:00
a3e794eb3b Tailscale autoconnect. 2023-07-08 19:17:40 +01:00
4866e6cf39 Update flake. 2023-07-08 06:38:49 +01:00
c06dec5414 Register service to consul. 2023-07-08 06:02:05 +01:00
0f7dc1cbe9 Add static webserver and storage for pi.paler.net. 2023-07-07 14:42:29 +01:00
734e8e0c61 Set immutable users in impermanence module. 2023-07-07 13:50:57 +01:00
dd971dcd1d Use a fixed uid/gid for storing files. 2023-07-07 13:50:40 +01:00