What is PromAlert?

Yet another Prometheus Alertmanager webhook processor inspired by qvl/promplot

How it works

Receive webhook from Alertmanager, draw images from alert expression, upload pictures to S3 bucket, generate public links, send a notification to Slack.

Why Promalert?

Standard prometheus slack receiver cant use threads and draw graphs from alert.

Main features:

• Threads to keep alert channel informative
• Updating init firing message, mark it as resolved
• Graph rendering with metric and alert level

Message logic

Post the new message in case when we can find a reference in the history example.

Otherwise, based on the status of the message, we apply the following logic:

Status = Firing

Post full message to thread with broadcasting. Set refiring footer to last known firing message. Store link to this message in memory (next resolving updates will edit this message) example.

Status = Resolved

Post short message (header + images) to thread example. Update footer of last fired message example.

Installation

Helm chart

Coming soon

Docker

Use env file to store required params

docker run -p 8080:8080 --env-file env.list kuzaxak/promalert

Configuration

The following tables list the configurable parameters of the PromAlert and their default values.

You can use a YAML configuration file or env variable. Package viper used to parse them. Environment variables prefix: PROMALERT_

Required params:

Parameter	Description	Env variable
slack_token	OAuth bot token	PROMALERT_SLACK_TOKEN
slack_channel	Slack channel to send	PROMALERT_SLACK_CHANNEL
prometheus_url	Prometheus URL	PROMALERT_PROMETHEUS_URL
s3_bucket	S3 bucket name	PROMALERT_S3_BUCKET
s3_region	S3 region	PROMALERT_S3_REGION

Additional params:

Parameter	Description	Default
http_port	HTTP port	8080
metric_resolution	Amount of point on the graph	100
debug	Verbose log output. Dump HTTP request to log	false
message_template	Slack message template. Go template syntax	config.example.yaml
header_template	Slack message header template. Go template syntax	config.example.yaml
footer_template	Slack message footer template. Go template syntax	config.example.yaml
graph_scale	Scale the graph image	1.0

AWS

AWS credentials parsed by aws-go-client in the following order:

1. Environment variables.
2. Shared credentials file.
3. If your application is running on an Amazon EC2 instance, IAM role for Amazon EC2.

Message templating

Template applies per alert in group. Data in the template . = Alert

Available functions:

Func	Arguments	Description	Example
toUpper	string	Format text to Uppercase	`{{ .Status
dateFormat	format string, time	Format date/time	{{ dateFormat "15:04:05" now }}
now		Retrieve current time	{{ now.String }}

Default message template:

  :chart_with_upwards_trend: *<{{ .GeneratorURL }}|Graph>*
  {{- if .Labels.runbook }} :notebook: *<{{ .Labels.runbook }}|Runbook>*{{ end }}
  {{- if .Annotations.runbook_url }} :notebook: *<{{ .Annotations.runbook_url }}|Runbook>*{{ end }}

  *Alert:* {{ if .Annotations.title }}{{ .Annotations.title }}{{ end }}{{ if .Annotations.summary }}{{ .Annotations.summary }}{{ end }}
  {{ if .Labels.severity }}*Severity:*  `{{ .Labels.severity }}`{{ end }}
  {{ if .Annotations.message }}*Message:*  {{ .Annotations.message }}{{ end }}
  {{ if .Annotations.description }}*Description:* {{ .Annotations.description }}{{ end }}
  *Details:*
    {{ range $key, $value := .Labels }} • {{ $key }}: {{ $value }}
    {{ end }}

Header template:

  *{{ .Labels.alertname }}*
  [Status]: {{if eq .Status "firing" }}:fire::fire::fire:{{else}}:white_check_mark::white_check_mark::white_check_mark:{{ end }}

Footer template:

  {{if eq .Status "firing" }}:fire: Refired{{else}}:white_check_mark: Resolved{{ end }} {{ dateFormat "15:04:05" now }}

Rendered message:

Build

docker build --rm -t promalert .

Bug Reporting

To submit a bug report use the GitHub bug tracker for the project:

Github Issues

License

GNU Lesser General Public License v3.0

See LICENSE to see the full text.