Service fails to start under systemd

[slarti5191]

Quotes in web.external-url are breaking the service with systemd.

Affected Puppet, Ruby, OS and module versions/distributions

• Puppet: 5.4.0
• Ruby: 2.0.0p648
• Distribution: CentOS 7
• Module version: 6.1.0

How to reproduce (e.g Puppet code you use)

Attempt to manage service with systemd and prometheus 2.0.0

What are you seeing

Service fails to start with this error:

Jul 31 13:53:14 prometheus.example.com prometheus[25352]: parse external URL "'/prometheus'": URL must not begin or end with quotes

Behavior is the same when external.url is not populated.

What behaviour did you expect instead

Normal service start

Output log

Any additional information you'd like to impart

Fixed with:

diff --git a/modules/prometheus/manifests/config.pp b/modules/prometheus/manifests/config.pp
index 9218a41..f198d6c 100644
--- a/modules/prometheus/manifests/config.pp
+++ b/modules/prometheus/manifests/config.pp
@@ -30,7 +30,7 @@ class prometheus::config {
         "--storage.tsdb.retention=${prometheus::server::storage_retention}",
         "--web.console.templates=${prometheus::server::shared_dir}/consoles",
         "--web.console.libraries=${prometheus::server::shared_dir}/console_libraries",
-        "--web.external-url='${prometheus::server::external_url}'",
+        "--web.external-url=${prometheus::server::external_url}",
       ]
     }

[bastelfreak]

Hey @slarti5191, thanks for rising this issue. Can you provide the puppet code you use to call the module? Do you explicitly set the $prometheus::server::external_url parameter?

[slarti5191]

Here is what I have - it is a near exact copy/paste from the example here

    manage_prometheus_server => true,
    version => '2.0.0',
    #lint:ignore:140chars
    alerts => { 'groups' => [{ 'name' => 'alert.rules', 'rules' => [{ 'alert' => 'InstanceDown', 'expr' => 'up == 0', 'for' => '5m', 'labels' => { 'severity' => 'page', }, 'annotations' => { 'summary' => 'Instance {{ $labels.instance }} down', 'description' => '{{ $labels.instance }} of job {{ $labels.job }} has been down for more than 5 minutes.' } }]}]},
    scrape_configs => [
      { 'job_name' => 'prometheus',
         'scrape_interval' => '10s',
         'scrape_timeout'  => '10s',
         'static_configs'  => [
           { 'targets' => [ 'localhost:9090' ],
             'labels'  => { 'alias'=> 'Prometheus'}
           }
         ]
       },
      { 'job_name' => 'files',
        'scrape_interval' => '5s',
        'scrape_timeout'  => '5s',
        'file_sd_configs'  => [
          { 'files' => [ '/etc/prometheus/conf.d/*.json' ],}
        ]
      }
    ],
    alertmanagers_config => [{ 'static_configs' => [{'targets' => [ 'localhost:9093' ]}]}],
  }

I tried specifying external_url => and encountered the same failure, once I removed the single quotes at line 33 of config.pp the issue went away.

[bastelfreak]

@slarti5191 We just merged #245, can you test if that fixes your issues?

[slarti5191]

Looks good - thank you!