Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

nixos/prometheus: remove services.prometheus.environmentFile #144984

Merged
Merged
Show file tree
Hide file tree
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 9 additions & 0 deletions nixos/doc/manual/from_md/release-notes/rl-2111.section.xml
Original file line number Diff line number Diff line change
Expand Up @@ -1803,6 +1803,15 @@ Superuser created successfully.
when its config file changes instead of restarting.
</para>
</listitem>
<listitem>
<para>
The option
<literal>services.prometheus.environmentFile</literal> has
been removed since it was causing
<link xlink:href="https://github.com/NixOS/nixpkgs/issues/126083">issues</link>
and Prometheus now has native support for secret files.
</para>
</listitem>
<listitem>
<para>
Dokuwiki now supports caddy! However
Expand Down
2 changes: 2 additions & 0 deletions nixos/doc/manual/release-notes/rl-2111.section.md
Original file line number Diff line number Diff line change
Expand Up @@ -508,6 +508,8 @@ In addition to numerous new and upgraded packages, this release has the followin

- A new option `services.prometheus.enableReload` has been added which can be enabled to reload the prometheus service when its config file changes instead of restarting.

- The option `services.prometheus.environmentFile` has been removed since it was causing [issues](https://github.com/NixOS/nixpkgs/issues/126083) and Prometheus now has native support for secret files.

- Dokuwiki now supports caddy! However
- the nginx option has been removed, in the new configuration, please use the `dokuwiki.webserver = "nginx"` instead.
- The "${hostname}" option has been deprecated, please use `dokuwiki.sites = [ "${hostname}" ]` instead
Expand Down
91 changes: 9 additions & 82 deletions nixos/modules/services/monitoring/prometheus/default.nix
Original file line number Diff line number Diff line change
Expand Up @@ -9,13 +9,6 @@ let

prometheusYmlOut = "${workingDir}/prometheus-substituted.yaml";

writeConfig = pkgs.writeShellScriptBin "write-prometheus-config" ''
PATH="${makeBinPath (with pkgs; [ coreutils envsubst ])}"
touch '${prometheusYmlOut}'
chmod 600 '${prometheusYmlOut}'
envsubst -o '${prometheusYmlOut}' -i '${prometheusYml}'
'';

triggerReload = pkgs.writeShellScriptBin "trigger-reload-prometheus" ''
PATH="${makeBinPath (with pkgs; [ systemd ])}"
if systemctl -q is-active prometheus.service; then
Expand Down Expand Up @@ -76,8 +69,8 @@ let
"--storage.tsdb.path=${workingDir}/data/"
"--config.file=${
if cfg.enableReload
then prometheusYmlOut
else "/run/prometheus/prometheus-substituted.yaml"
then "/etc/prometheus/prometheus.yaml"
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also how does the new secrets work? Can we add documentation for this?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I added the alternatives to the release notes and to the mkRemovedOptionModule error message.

else prometheusYml
}"
"--web.listen-address=${cfg.listenAddress}:${builtins.toString cfg.port}"
"--alertmanager.notification-queue-capacity=${toString cfg.alertmanagerNotificationQueueCapacity}"
Expand Down Expand Up @@ -1625,51 +1618,6 @@ in
(<literal>switch-to-configuration</literal>) that changes the prometheus
configuration only finishes successully when prometheus has finished
loading the new configuration.

Note that prometheus will also get reloaded when the location of the
<option>environmentFile</option> changes but not when its contents
changes. So when you change it contents make sure to reload prometheus
manually or include the hash of <option>environmentFile</option> in its
name.
'';
};

environmentFile = mkOption {
type = types.nullOr types.path;
default = null;
example = "/root/prometheus.env";
description = ''
Environment file as defined in <citerefentry>
<refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum>
</citerefentry>.

Secrets may be passed to the service without adding them to the
world-readable Nix store, by specifying placeholder variables as
the option value in Nix and setting these variables accordingly in the
environment file.

Environment variables from this file will be interpolated into the
config file using envsubst with this syntax:
<literal>$ENVIRONMENT ''${VARIABLE}</literal>

<programlisting>
# Example scrape config entry handling an OAuth bearer token
{
job_name = "home_assistant";
metrics_path = "/api/prometheus";
scheme = "https";
bearer_token = "\''${HOME_ASSISTANT_BEARER_TOKEN}";
[...]
}
</programlisting>

<programlisting>
# Content of the environment file
HOME_ASSISTANT_BEARER_TOKEN=someoauthbearertoken
</programlisting>

Note that this file needs to be available on the host on which
<literal>Prometheus</literal> is running.
'';
};

Expand Down Expand Up @@ -1830,40 +1778,26 @@ in
uid = config.ids.uids.prometheus;
group = "prometheus";
};
environment.etc."prometheus/prometheus.yaml" = mkIf cfg.enableReload {
source = prometheusYml;
};
systemd.services.prometheus = {
wantedBy = [ "multi-user.target" ];
after = [ "network.target" ];
preStart = mkIf (!cfg.enableReload) ''
${lib.getBin pkgs.envsubst}/bin/envsubst -o "/run/prometheus/prometheus-substituted.yaml" \
-i "${prometheusYml}"
'';
serviceConfig = {
ExecStart = "${cfg.package}/bin/prometheus" +
optionalString (length cmdlineArgs != 0) (" \\\n " +
concatStringsSep " \\\n " cmdlineArgs);
ExecReload = mkIf cfg.enableReload "+${reload}/bin/reload-prometheus";
User = "prometheus";
Restart = "always";
EnvironmentFile = mkIf (cfg.environmentFile != null && !cfg.enableReload) [ cfg.environmentFile ];
Copy link
Member

@Mic92 Mic92 Nov 7, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should there maybe a deprecation phase with a warning?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Since this is causing issues (#126083, #127889 (comment)) and since there's now an alternative I think it's better to go without a deprecation cycle. I did now add a mkRemovedOptionModule so that users get a helpful error message when they define services.prometheus.environmentFile.

RuntimeDirectory = "prometheus";
RuntimeDirectoryMode = "0700";
WorkingDirectory = workingDir;
StateDirectory = cfg.stateDir;
StateDirectoryMode = "0700";
};
};
systemd.services.prometheus-config-write = mkIf cfg.enableReload {
wantedBy = [ "prometheus.service" ];
before = [ "prometheus.service" ];
serviceConfig = {
Type = "oneshot";
User = "prometheus";
StateDirectory = cfg.stateDir;
StateDirectoryMode = "0700";
EnvironmentFile = mkIf (cfg.environmentFile != null) [ cfg.environmentFile ];
ExecStart = "${writeConfig}/bin/write-prometheus-config";
};
};
# prometheus-config-reload will activate after prometheus. However, what we
# don't want is that on startup it immediately reloads prometheus because
# prometheus itself might have just started.
Expand All @@ -1873,26 +1807,19 @@ in
# harmless message and then stay active (RemainAfterExit).
#
# Then, when the config file has changed, switch-to-configuration notices
# that this service has changed and needs to be reloaded
# (reloadIfChanged). The reload command then actually writes the new config
# and reloads prometheus.
# that this service has changed (restartTriggers) and needs to be reloaded
# (reloadIfChanged). The reload command then reloads prometheus.
systemd.services.prometheus-config-reload = mkIf cfg.enableReload {
wantedBy = [ "prometheus.service" ];
after = [ "prometheus.service" ];
reloadIfChanged = true;
restartTriggers = [ prometheusYml ];
serviceConfig = {
Type = "oneshot";
User = "prometheus";
StateDirectory = cfg.stateDir;
StateDirectoryMode = "0700";
EnvironmentFile = mkIf (cfg.environmentFile != null) [ cfg.environmentFile ];
RemainAfterExit = true;
TimeoutSec = 60;
ExecStart = "${pkgs.logger}/bin/logger 'prometheus-config-reload will only reload prometheus when reloaded itself.'";
ExecReload = [
"${writeConfig}/bin/write-prometheus-config"
"+${triggerReload}/bin/trigger-reload-prometheus"
];
ExecReload = [ "${triggerReload}/bin/trigger-reload-prometheus" ];
};
};
};
Expand Down
11 changes: 1 addition & 10 deletions nixos/tests/prometheus.nix
Original file line number Diff line number Diff line change
Expand Up @@ -130,14 +130,10 @@ in import ./make-test-python.nix {

# This configuration just adds a new prometheus job
# to scrape the node_exporter metrics of the s3 machine.
# We also use an environmentFile to test if that works correctly.
services.prometheus = {
environmentFile = pkgs.writeText "prometheus-config-env-file" ''
JOB_NAME=s3-node_exporter
'';
scrapeConfigs = [
{
job_name = "$JOB_NAME";
job_name = "s3-node_exporter";
static_configs = [
{
targets = [ "s3:9100" ];
Expand Down Expand Up @@ -232,11 +228,6 @@ in import ./make-test-python.nix {
# Check if prometheus responds to requests:
prometheus.wait_for_unit("prometheus.service")

# Check if prometheus' config file is correctly locked down because it could contain secrets.
prometheus.succeed(
"stat -c '%a %U' /var/lib/prometheus2/prometheus-substituted.yaml | grep '600 prometheus'"
)

prometheus.wait_for_open_port(${toString queryPort})
prometheus.succeed("curl -sf http://127.0.0.1:${toString queryPort}/metrics")

Expand Down