infrastructure/modules/nixos/dgn-monitoring/default.nix

99 lines
2.2 KiB
Nix

# SPDX-FileCopyrightText: 2024 Tom Hubrecht <tom.hubrecht@dgnum.eu>
#
# SPDX-License-Identifier: EUPL-1.2
{
config,
lib,
pkgs,
meta,
name,
nodeMeta,
...
}:
let
inherit (lib)
filterAttrs
mapAttrs
mapAttrsToList
mkDefault
mkEnableOption
mkForce
mkIf
mkOption
;
inherit (lib.types) attrsOf;
cfg = config.dgn-monitoring;
in
{
imports = [ ./exporters.nix ];
options.dgn-monitoring = {
enable = mkEnableOption "the DGNum monitoring system" // {
default = true;
};
scrapeConfigs = mkOption {
type = attrsOf (pkgs.formats.yaml { }).type;
description = ''
Specifications of `scrape_config` sections.
'';
};
};
config = mkIf cfg.enable {
dgn-monitoring.scrapeConfigs =
mapAttrs
(_: cfg: {
static_configs = mkDefault [ { targets = [ "127.0.0.1:${builtins.toString cfg.port}" ]; } ];
})
(
filterAttrs (
name: cfg:
!(builtins.elem name [
"assertions"
"warnings"
"blackbox"
"unifi-poller"
"domain"
"minio"
"idrac"
"pve"
"tor"
])
&& cfg.enable
) config.services.prometheus.exporters
);
services.vmagent = {
enable = true;
flags = {
"remoteWrite.url" = "http://${meta.network.storage01.netbirdIp}:8428/api/v1/write";
"remoteWrite.label" = "node=${name}";
};
prometheusConfig = {
scrape_configs = mapAttrsToList (job_name: value: value // { inherit job_name; }) cfg.scrapeConfigs;
global = {
scrape_interval = "15s";
external_labels.hostname = "${name}.${nodeMeta.site}.infra.dgnum.eu";
};
};
};
services.journald.upload = {
enable = true;
settings = {
Upload.URL = "http://${meta.network.storage01.netbirdIp}:9428/insert/journald";
};
};
# Don't restart too often to reduce e-mail notifications when the network or the database is down
systemd.services.systemd-journal-upload.serviceConfig.RestartSec = mkForce 60;
};
}