Data

Tools

Indexes

Applications

Experiments

Awesome

An open API service indexing awesome lists of open source software.

https://github.com/orange-cloudfoundry/statusetat

status page with very HA in mind and notifier
https://github.com/orange-cloudfoundry/statusetat

high-availability notifications status-page

Last synced: 5 months ago
JSON representation

status page with very HA in mind and notifier

Awesome Lists containing this project

README 

          
# Statusetat [![ginkgo test badge](https://github.com/orange-cloudfoundry/statusetat/workflows/ginkgo-test/badge.svg)](https://github.com/orange-cloudfoundry/statusetat/actions?query=workflow%3Aginkgo-test)



status page with very HA in mind and notification system to send information about incident(s) or scheduled task(s) 

to external system like slack, email or whatever plugin you implement.



[![home one incident and maintenance](/screenshots/home_one_incident_and_maintenance.png)](/screenshots/home_one_incident_and_maintenance.png)



Features:

- Easy admin web interface to manage incidents and scheduled tasks

- Send notifications to external system (slack, email, plugins...) when incident come

- Full rest api to create/read/update/delete incident/scheduled tasks

- Subscribes systems



## Why another status page ?



A status page **must** be more resilient than the service you want to be monitored because you need to have the page 

always available even when your a part of your infra is partially down to give the information to your client.



When you're a hosting service be more resilient on your status page become complicated. 

This implementation allow you to set multiple storage system in same time to ensure you always can store/retrieve incidents and 

scheduled tasks.



Statusetat will replicate in all storage what you write and ensure that the data can be retrieved.



## Getting started



1. Download latest release for your system

2. Create a `config.yml` file with this content for a dev deployment



```yaml

listen: 0.0.0.0:8080 # this will listen in http on port 8080

targets:

  - sqlite://./store.db

log:

  level: debug

components:

  - name: my-service

    description: description of my service

    group: ~

notifiers:

- type: slack

  params:

    endpoint: http://my-mattermost.com/hooks/aj6na7hz5ib1mfs3eggc1bu8tk

- type: plugin

  params:

    path: /path/to/my/plugin.so

username: admin

password: admin

cookie_key: encryption-key-for-cookies

base_info:

  base_url: http://localhost:8080

  support: https://doc.to.my.service.com

  contact: http://my-mattermost.com/channel/town

  time_zone: UTC

```



## Configuration



For understanding config definition format:

- `[]` means optional (by default parameter is required)

- `<>` means type to use



### Root configuration in config.yml



```yaml

# Listen address for listening in http

# If empty and env var `PORT` set default listen will 0.0.0.0:${PORT}

[ listen:  | default = 0.0.0.0:8080 ]



# If config tls not set no https will be enabled

# only set tls if you want https on statusetat

tls:

  # Path to cert file

  [ cert_file:  ]

  # Path to key file

  [ key_file:  ]



log:

  # log level to use for server

  # you can chose: `trace`, `debug`, `info`, `warn`, `error`, `fatal` or `panic`

  [ level:  | default = info ]

  # Set to true to force not have color when seeing logs

  [ no_color:  ]

  # et to true to see logs as json format

  [ in_json:  ]

# username for basic authentication to access admin page or api

username: 

# password for basic authentication to access admin page or api

password: 

# cookie key for cookie encryption (generate a random value and set it here)

cookie_key: 

base_info:

  # Url of your statusetat for forging urls on notifications

  base_url: 

  # Title of your status page

  [ title:  | default = "Statusetat" ]

  # Url for your user to know where they can found support/doc page

  [ support:  ]

  # Url for your user to know where they can contact you

  [ contact:  ]

  # Default timezone when creating/updating/show incident(s)

  [ timezone:  | default = "UTC" ]

# customize web page

theme:

  # Markdown content to show before status boxes

  [ pre_status:  ]

  # Markdown content to show after status boxes

  [ post_status:  ]

  # Markdown content to show before timeline

  [ pre_timeline:  ]

  # Markdown content to show after timeline

  [ post_timeline:  ]

  # Markdown content to show before maintenance/scheduled tasks box

  [ pre_maintenance:  ]

  # Markdown content to show after maintenance/scheduled tasks box

  [ post_maintenance:  ]

  # Markdown content to put in footer

  [ footer:  ]

  # Persistent name to display, default to Persistent Incident but you could set Known Issues for example

  [ persistent_display_name:  ]

  # Markdown content to show before persistent incident tasks box

  [ pre_persistent:  ]

  # Markdown content to show after persistent incident tasks box

  [ post_persistent:  ]

# list of targets store to use, this can be in form of:

# mysql://user:password@host:3306?options

# mariadb://user:password@host:3306?options

# postgres://user:password@host:3306?options

# sqlite://:memory:

# sqlite:///a/path

# s3://access_key_id:access_key_secret@host.com/bucket?region=us-east-1&insecure-skip-verify=false

# file:///a/path/to/a/folder

targets:

- 

notifiers:

[ -  ]

```



### notifiers configuration



```yaml

# type of notifier to use, you can found them in notifiers section of this doc

type: 

# only trigger for particular component(s)

[ for:  ]

# map of params for the notifier you use

params:

  [ :  ]

```



### for_component configurations



```yaml

# Set to true to mandatory match all filter

# Think that you want to use an AND condition on true and OR condition on false

[require_all:  | default = false ]

# Group which match component(s)

groups:

[ -  ]

# Names which match component(s)

names:

[ -  ]

```



## Notifiers



### Slack



Notify on a slack channel or mattermost channel new incidents or scheduled tasks.



- **Type name**: `slack`

- **Params**:



```yaml

# incoming endpoint url

endpoint: 

# Specify channel to send notif

[ channel:  ]

# Specify username to show on message

[ channel:  ]

# Specify icon emoji incident to show on message

[ icon_emoji_incident:  | default = "bell" ]

# Specify icon emoji scheduled tasks to show when planified

[ icon_emoji_scheduled:  | default = "clock1" ]

# Specify icon emoji scheduled tasks to show when started

[ icon_emoji_scheduled_started:  | default = "clock1" ]

# Specify icon emoji scheduled tasks to show when finished

[ icon_emoji_scheduled_finish:  | default = "clock1" ]

# skip ssl verification when sending notification

[ insecure_skip_verify:  ]

```



### Grafana annotation



Put a grafana annotation when incident start and when it's finished on a graph.



- **Type name**: `grafana_annotation`

- **Params**:



```yaml

# Grafana api endpoint

endpoint: 

# Grafana api key to use

api_key: 

# Dashboard id to set annotation

dashboard_id: 

# Panel id in dashboard to set annotation

panel_id: 

# tag used in grafana to filter global annotations

[ tag:  ]

# time zone to use when sending annotation

[ time_zone:  ]

# skip ssl verification when sending notification

[ insecure_skip_verify:  ]

# organization id to use when sending annotation

[ org_id:  ]

```



### Email



Send incidents and scheduled tasks to subscribers via subscribe api and/or emails in notifier param.

**Type name**: `email`

**Params**:



```yaml

# Email server host

host: 

# Email server port

[ port:  | default = 25 ]

# Username to use when connect to email server

[ username:  ]

# Password to use when connect to email server

[ password:  ]

# Set to true to use ssl/tls when connect to email server

[ use_ssl:  ]

# Email to insert in from field

[ From:  | default = "no-reply@local" ]

# List of emails address to send email, e.g.: admins of the monitored service

subscribers:

[ -  ]

# Go template to use to write subject for incident

[ subject_incident:  | default = "[{{ .TitleSite }} {{ .Incident.State | textIncidentState | title }} Incident] {{ .IncidentTitle | title }}" ]

# Go template to use to write subject for scheduled tasks

[ subject_scheduled:  | default = "[{{ .TitleSite }} Scheduled task] {{ .IncidentTitle | title }}" ]

# Go template to use to write content of email when get an incident

[ txt_incident:  ]

```



For customize content of email you can see data passed to template at https://github.com/orange-cloudfoundry/statusetat/blob/master/notifiers/email/email.go#L223-L236 

and template function available at https://github.com/orange-cloudfoundry/statusetat/blob/master/extemplate/template.go#L25-L51



## Plugin



You can create a plugin for notification for your own need.



Plugin are grpc golang plugin which doc can be found here: https://github.com/hashicorp/go-plugin .



You must implement [plugin.Notifier](/notifiers/plugin/interface.go).



Example of implementation:



```go

package main



import (

	pluginhc "github.com/hashicorp/go-plugin"

	"github.com/orange-cloudfoundry/statusetat/config"

	"github.com/orange-cloudfoundry/statusetat/models"

	"github.com/orange-cloudfoundry/statusetat/notifiers/plugin"

	log "github.com/sirupsen/logrus"

	"strings"

)



type LogNotifier struct {

}



func (n LogNotifier) PreCheck(incident *models.Incident) error {

	return nil

}



func (n *LogNotifier) Init(baseInfo config.BaseInfo, params map[string]interface{}) error {

	return nil

}



func (n LogNotifier) Name() (string, error) {

	return "log_notifier", nil

}



func (n LogNotifier) Id() (string, error) {

	return "log_notifier1", nil

}



func (n LogNotifier) MetadataFields() ([]models.MetadataField, error) {

	return []models.MetadataField{}, nil

}



func (n LogNotifier) Notify(notifyReq *models.NotifyRequest) error {

	log.

		WithField("subscribers", strings.Join(notifyReq.Subscribers, ", ")).

		WithField("triggerred_by_user", notifyReq.TriggerByUser).

		Info(notifyReq.Incident)

	return nil

}



func main() {

	pluginhc.Serve(&pluginhc.ServeConfig{

		HandshakeConfig: plugin.Handshake,

		Plugins: map[string]pluginhc.Plugin{

			"notifier": &plugin.NotifierGRPCPlugin{

				Impl: &LogNotifier{},

			},

		},



		// A non-nil value here enables gRPC serving for this plugin...

		GRPCServer: pluginhc.DefaultGRPCServer,

	})

}

```



You can build it with command line `go build -o log-notif .`



You can now use it in your config with:



- **Type name**: `plugin` or notifier name you have set in func signature `Name() string`

- **Params**:

```yaml

# path to .so to load

# in our example it would be `./log-notif`

path: 

# all remaining arguments will be pass at Init

```



## Api 



To document



## Credits



This project was heavily inspired by [statusfy](https://github.com/juliomrqz/statusfy) mostly on the design part and 

models of incidents/components.



[Cachethq](https://cachethq.io/) is another source of inspiration for some details on the admin design part.