diff options
| -rw-r--r-- | krebs/2configs/shack/prometheus/alert-rules.nix | 12 | ||||
| -rw-r--r-- | krebs/2configs/shack/prometheus/alertmanager-telegram.nix | 4 | ||||
| -rw-r--r-- | krebs/2configs/shack/prometheus/templates/shack.tmpl | 25 |
3 files changed, 33 insertions, 8 deletions
diff --git a/krebs/2configs/shack/prometheus/alert-rules.nix b/krebs/2configs/shack/prometheus/alert-rules.nix index 730921bee..1c2d0b1ad 100644 --- a/krebs/2configs/shack/prometheus/alert-rules.nix +++ b/krebs/2configs/shack/prometheus/alert-rules.nix @@ -10,28 +10,28 @@ in { { alert = "RootPartitionFull"; for = "30m"; - expr = ''(node_filesystem_avail_bytes{alias="wolf",mountpoint="/"} * 100) / node_filesystem_size_bytes{alias="wolf",mountpoint="/"} < ${disk_free_threshold}''; + expr = ''(node_filesystem_avail_bytes{alias="wolf.shack",mountpoint="/"} * 100) / node_filesystem_size_bytes{alias="wolf.shack",mountpoint="/"} < ${disk_free_threshold}''; labels.severity = "warning"; annotations.summary = "{{ $labels.alias }} root disk full"; - annotations.url = "http://grafana.shack/"; + annotations.url = "http://grafana.shack/d/hb7fSE0Zz/shack-system-dashboard?orgId=1&var-job=node&var-hostname=All&var-node=wolf.shack:9100&var-device=All&var-maxmount=%2F&var-show_hostname=wolf"; annotations.description = ''The root disk of {{ $labels.alias }} has {{ $value | printf "%.2f" }}% free disk space (Threshold at ${disk_free_threshold}%).A vast number of shackspace services will stop working. CI for deploying new configuration will also seize working. Log in to the system and run `nix-collect-garbage -d` and clean up the shack share folder in `/home/share` .If this does not help you can check `du -hs /var/ | sort -h`, run `docker system prune` or if you are really desperate run `du -hs / | sort -h` and go through the folders recursively until you've found something to delete''; } { alert = "RootPartitionFull"; for = "30m"; - expr = ''(node_filesystem_avail_bytes{alias="puyak",mountpoint="/"} * 100) / node_filesystem_size_bytes{alias="puyak",mountpoint="/"} < ${disk_free_threshold}''; + expr = ''(node_filesystem_avail_bytes{alias="puyak.shack",mountpoint="/"} * 100) / node_filesystem_size_bytes{alias="puyak.shack",mountpoint="/"} < ${disk_free_threshold}''; labels.severity = "warning"; annotations.summary = "{{ $labels.alias }} root disk full"; - annotations.url = "http://grafana.shack/"; + annotations.url = "http://grafana.shack/d/hb7fSE0Zz/shack-system-dashboard?orgId=1&var-job=node&var-hostname=All&var-node=wolf.shack:9100&var-device=All&var-maxmount=%2F&var-show_hostname=puyak"; annotations.description = ''The root disk of {{ $labels.alias }} has {{ $value | printf "%.2f" }}% free disk space (Threshold at ${disk_free_threshold}%).Prometheus will not be able to create new alerts and CI for deploying new configuration will also seize working. Log in to the system and run `nix-collect-garbage -d` and if this does not help you can check `du -hs /var/ | sort -h`, run `docker system prune` or if you are really desperate run `du -hs / | sort -h` and go through the folders recursively until you've found something to delete''; } { alert = "HostDown"; - expr = ''up{alias="wolf"} == 0''; + expr = ''up{alias="wolf.shack"} == 0''; for = "5m"; labels.severity = "page"; annotations.summary = "Instance {{ $labels.alias }} down for 5 minutes"; - annotations.url = "http://grafana.shack/"; + annotations.url = "http://grafana.shack/d/hb7fSE0Zz/shack-system-dashboard?orgId=1&var-job=node&var-hostname=All&var-node=wolf.shack:9100&var-device=All&var-maxmount=%2F&var-show_hostname=wolf"; annotations.description = ''Host {{ $labels.alias }} went down and has not been reconnected after 5 minutes. This is probably bad news, try to restart the host via naproxen ( http://naproxen.shack:8006 ). Wolf being down means that CI,glados automation, light management and a couple of other services will not work anymore.''; } ]; diff --git a/krebs/2configs/shack/prometheus/alertmanager-telegram.nix b/krebs/2configs/shack/prometheus/alertmanager-telegram.nix index 9d0ef45e8..2df63e520 100644 --- a/krebs/2configs/shack/prometheus/alertmanager-telegram.nix +++ b/krebs/2configs/shack/prometheus/alertmanager-telegram.nix @@ -8,10 +8,10 @@ DynamicUser = true; StateDirectory = "alertbot"; ExecStart = ''${pkgs.alertmanager-bot-telegram}/bin/alertmanager-bot \ - --alertmanager.url=http://alert.prometheus.shack --log.level=info \ + --alertmanager.url=http://alert.prometheus.shack --log.level=debug \ --store=bolt --bolt.path=/var/lib/alertbot/bot.db \ --listen.addr="0.0.0.0:16320" \ - --template.paths=${pkgs.alertmanager-bot-telegram}/templates/default.tmpl''; + --template.paths=${./templates}/shack.tmpl''; }; }; } diff --git a/krebs/2configs/shack/prometheus/templates/shack.tmpl b/krebs/2configs/shack/prometheus/templates/shack.tmpl new file mode 100644 index 000000000..9295f019f --- /dev/null +++ b/krebs/2configs/shack/prometheus/templates/shack.tmpl @@ -0,0 +1,25 @@ +{{ define "telegram.default" }} +{{range .Alerts -}} +{{ $severity := index .Labels "severity" }} +{{ $desc := "No Description" }} +{{ if eq .Status "firing" }} + {{ $desc = index .Annotations "description" }} + {{- if eq $severity "critical" -}} + <i><u><b>[CRITICAL]</b></u></i> + {{- else if eq $severity "warning" -}} + <u><b>[WARNING]</b></u> + {{- else -}} + <b>[{{ $severity }}]</b> + {{- end -}} +{{ else -}} + {{ $desc = "The issue has been resolved" }} + <del>[RESOLVED]</del> +{{- end }} {{ index .Labels "alertname"}}: {{ index .Annotations "summary"}} + +{{ $desc }} + +Alert Links: +* <a href="{{ index .Annotations "url"}}">Grafana</a> +* <a href="{{ .GeneratorURL }}">Source</a> +{{end -}} +{{end}} |