alerta-server 4.2.1

Alerta server WSGI application

Introduction

Alerta is a monitoring tool that allows alerts from many different systems to be consolidated into a single view.

Currently there are integrations for tools that support standard protocols such as SNMP, syslog and HTTP. There are also specific integrations for popular monitoiring tools such as Nagios, Zabbix and Riemann.

Installation

Installing this package makes available a alert-sender and alert-query tool that can be used to send alerts to the alerta system and to query the alert database:

$ pip install alerta

Configuration

For a basic configuration that can be used to test the client tools against a demo alerta server, use:

[DEFAULT]
timezone = Europe/London
api_host = api.alerta.io
api_port = 80

[alert-query]
colour = yes

Copy this configuration to /etc/alerta/alerta.conf or $HOME/.alerta.conf files:

$ mv /path/to/alerta.conf.sample $HOME/.alerta.conf

or use the ALERTA_CONF environment variable:

$ export ALERTA_CONF=/path/to/alerta.conf

Basic Usage

Abbreviated usage for both commands is shown below:

usage: alert-sender [-r RESOURCE] [-e EVENT] [-C CORRELATE] [-g GROUP]
                    [-v VALUE] [--status STATUS] [-s SEVERITY] [-E ENV]
                    [-S SERVICE] [-T TAGS] [-t TEXT] [--summary TEXT]
                    [--more TEXT] [--graphs URLS] [-o TIMEOUT]
                    [--type EVENT_TYPE] [-H] [-O ORIGIN] [-q] [-d]


usage: alert-query [-h] [-c FILE] [--minutes MINUTES] [--hours HOURS]
                   [--days DAYS] [-i ALERTID] [-E ENV] [--not-environment ENV]
                   [-S SERVICE] [--not-service SERVICE] [-r RESOURCE]
                   [--not-resource RESOURCE] [-s SEVERITY]
                   [--not-severity SEVERITY] [--status STATUS]
                   [--not-status STATUS] [-e EVENT] [--not-event EVENT]
                   [-g GROUP] [--not-group GROUP] [--origin ORIGIN]
                   [--not-origin ORIGIN] [-v VALUE] [--not-value VALUE]
                   [-T TAGS] [--not-tags TAGS] [-t TEXT] [--not-text TEXT]
                   [--type EVENT_TYPE] [--not-type TYPE]
                   [--repeat {true,false}] [--show SHOW] [--oneline]
                   [--date DATE] [--format FORMAT] [-o SORTBY] [-w]
                   [-n INTERVAL] [--count LIMIT] [-q QUERY] [--no-header]
                   [--no-footer] [--color] [--output OUTPUT] [-j] [-X] [-d]
                   [--version] [--debug] [--verbose] [--log-dir DIR]
                   [--log-file FILE] [--pid-dir DIR] [--use-syslog]
                   [--use-stderr] [--yaml-config FILE] [--show-settings]

Examples

To send a DiskFull warning alert for /tmp on myhost, use:

$ alert-sender --resource myhost:/tmp --event DiskFull --severity warning

To list all alerts for myhost, use:

$ alert-query --resource myhost

Trouble-shooting

To use curl to request the same URL for a query, use:

$ alert-query --dry-run | sh

And for an alert send, use:

$ alert-sender -r myhost -e test --dry-run | sh

Python API

A python API can be used to generate alerts:

>>> from alerta.common.api import ApiClient
>>> from alerta.common.alert import Alert
>>>
>>> client = ApiClient(host='api.alerta.io', port=80)
>>> alert = Alert(resource="foo", event="bar")
>>>
>>> client.send(alert)
u'8e9c4736-c2a8-4b4d-8638-07dad6ed1d2b'
>>>

The python API can also be used to query for alerts:

>>> from alerta.common.api import ApiClient
>>>
>>> client = ApiClient(host='api.alerta.io', port=80)
>>> r = client.query()
>>> r['status']
u'ok'
>>> pp = pprint.PrettyPrinter(indent=4)
>>> pp.pprint(r['alerts']['severityCounts'])
{   u'cleared': 0,
    u'critical': 1,
    u'debug': 0,
    u'indeterminate': 0,
    u'informational': 0,
    u'major': 2,
    u'minor': 1,
    u'normal': 4,
    u'security': 0,
    u'unknown': 0,
    u'warning': 1}