30 KiB
Icinga 2 API
Introduction
The Icinga 2 API allows you to manage configuration objects and resources in a simple, programmatic way using HTTP requests.
The endpoints are logically separated allowing you to easily make calls to
- run actions (reschedule checks, etc.)
- query, create, modify and delete config objects
- create/update/delete configuration objects
- manage configuration packages
- subscribe to event streams
This chapter will start with a general overview followed by detailed information about specific endpoints.
Requests
Any tool capable of making HTTP requests can communicate with the API, for example curl.
Requests are only allowed to use the HTTPS protocol so that traffic remains encrypted.
By default the Icinga 2 API listens on port 5665
sharing this
port with the cluster communication protocol. This can be changed
by setting the bind_port
attribute in the ApiListener
configuration object in the /etc/icinga2/features-available/api.conf
file.
Supported request methods:
Method | Usage |
---|---|
GET | Retrieve information about configuration objects. Any request using the GET method is read-only and does not affect any objects. |
POST | Update attributes of a specified configuration object. |
PUT | Create a new object. The PUT request must include all attributes required to create a new object. |
DELETE | Remove an object created by the API. The DELETE method is idempotent and does not require any check if the object actually exists. |
HTTP Statuses
The API will return standard HTTP statuses including error codes.
When an error occurs, the response body will contain additional information about the problem and its source.
A status in the range of 200 generally means that the request was succesful and no error was encountered.
Return codes within the 400 range indicate that there was a problem with the request. Either you did not authenticate correctly, you are missing the authorization for your requested action, the requested object does not exist or the request was malformed.
A status in the range of 500 generally means that there was a server-side problem and Icinga 2 is unable to process your request currently.
Ask your Icinga 2 system administrator to check the icinga2.log
file for further
troubleshooting.
Responses
Succesful requests will send back a response body containing a results
list. Depending on the number of affected objects in your request, the
results may contain one or more entries.
The output will be sent back as JSON object:
{
"results": [
{
"code": 200.0,
"status": "Object was created."
}
]
}
Authentication
There are two different ways for authenticating against the Icinga 2 API:
- username and password using HTTP basic auth
- X.509 certificate with client CN
In order to configure a new API user you'll need to add a new ApiUser
configuration object. In this example root
will be the basic auth username
and the password
attribute contains the basic auth password.
vim /etc/icinga2/conf.d/api-users.conf
object ApiUser "root" {
password = icinga"
}
Alternatively you can use X.509 client certificates by specifying the client_cn
the API should trust.
vim /etc/icinga2/conf.d/api-users.conf
object ApiUser "api-clientcn" {
password = "CertificateCommonName"
}
An ApiUser
object can have both methods configured. Sensitive information
such as the password will not be exposed through the API itself.
New installations of Icinga 2 will automatically generate a new ApiUser
named root
with a generated password in the /etc/icinga2/conf.d/api-users.conf
file.
You can manually invoke the cli command icinga2 api setup
which will generate
a new local CA, self-signed certificate and a new API user configuration.
Once the API user is configured make sure to restart Icinga 2:
# service icinga2 restart
Now pass the basic auth information to curl and send a GET request to the API:
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/status'
In case you will get an error message make sure to check the API user credentials.
Permissions
By default an api user does not have any permissions to perform actions on the url endpoints.
Permissions for api users must be specified in the permissions
attribute
as array. The array items can be a list of permission strings with wildcard
matches.
Example for an api user with all permissions:
permissions = [ "*" ]
A yet more sophisticated approach is to specify additional permissions and their filters. The latter must be defined as lamdba function returning a boolean expression.
The permission
attribute contains the action and the specific capitalized
object type name. Instead of the type name it is also possible to use a wildcard
match.
The following example allows the api user to query all hosts and services with
the custom host attribute os
matching the regular expression ^Linux
.
permissions = [
{
permission = "objects/query/Host"
filter = {{ regex("^Linux", host.vars.os) }}
},
{
permission = "objects/query/Service"
filter = {{ regex("^Linux", host.vars.os) }}
},
]
Available permissions for specific url endpoints:
Permissions | Url Endpoint |
---|---|
actions/;<action;> | /v1/actions |
config/query | /v1/config |
config/modify | /v1/config |
objects/query/;<type;> | /v1/objects |
objects/create/;<type;> | /v1/objects |
objects/modify`/;<type;> | /v1/objects |
objects/delete/;<type;> | /v1/objects |
status/query | /v1/status |
The required actions or types can be replaced by using a wildcard match ("*").
Parameters
Depending on the request method there are two ways of passing parameters to the request:
- JSON body (
POST
,PUT
) - Query string (
GET
,DELETE
)
Reserved characters by the HTTP protocol must be passed url-encoded as query string, e.g. a
space becomes %20
.
Example for query string:
/v1/objects/hosts?filter=match(%22nbmif*%22,host.name)&attrs=host.name&attrs=host.state
Example for JSON body:
{ "attrs": { "address": "8.8.4.4", "vars.os" : "Windows" } }
TODO
Filters
Use the same syntax as for apply rule expressions for filtering specific objects.
Example for all services in NOT-OK state:
https://localhost:5665/v1/objects/services?filter=service.state!=0
Example for matching all hosts by name (Note: "
are url-encoded as %22
):
https://localhost:5665/v1/objects/hosts?filter=match(%22nbmif*%22,host.name)
TODO
Output Format
The request and reponse body contain a JSON encoded string.
Version
Each url contains the version string as prefix (currently "/v1").
Url Endpoints
The Icinga 2 API provides multiple url endpoints:
Url Endpoints | Description |
---|---|
/v1/actions | Endpoint for running specific API actions. |
/v1/config | Endpoint for managing configuration modules. |
/v1/events | Endpoint for subscribing to API events. |
/v1/objects | Endpoint for querying, creating, modifying and deleting config objects. |
/v1/status | Endpoint for receiving icinga2 status and statistics. |
/v1/types | Endpoint for listing Icinga 2 configuration object types and their attributes. |
Please check the respective sections for detailed urls and parameters.
Actions
There are several actions available for Icinga 2 provided by the actions
url endpoint.
In case you have been using the external commands in the past, the API actions provide a yet more powerful interface with filters and even more functionality.
Actions require specific target types (e.g. type=Host
) and a filter expression.
TODO Figure out the final names.
Action name | Parameters | Target types | Notes |
---|---|---|---|
process-check-result | exit_status; plugin_output; check_source; performance_data[]; check_command[]; execution_end; execution_start; schedule_end; schedule_start | Service; Host | - |
reschedule-check | {next_check}; {(force_check)} | Service; Host | - |
acknowledge-problem | author; comment; {timestamp}; {(sticky)}; {(notify)} | Service; Host | - |
remove-acknowledgement | - | Service; Host | - |
add-comment | author; comment | Service; Host | - |
remove-comment | - | Service;Host | - |
remove-comment-by-id | comment_id | - | - |
delay-notifications | timestamp | Service;Host | - |
add-downtime | start_time; end_time; duration; author; comment; {trigger_id}; {(fixed)} | Service; Host; ServiceGroup; HostGroup | Downtime for all services on host x? |
remove-downtime | - | Service; Host | - |
remove-downtime-by-id | downtime_id | - | - |
send-custom-notification | options[]; author; comment | Service; Host | - |
enable-passive-checks | - | Service; Host; ServiceGroup; HostGroup | "System" as target? disable-passive-checks | - | Service; Host; ServiceGroup; HostGroup | diable all passive checks for services of hosts y in hostgroup x? enable-active-checks | - | Host; HostGroup | - disable-active-checks | - | Host; HostGroup | - enable-notifications | - | Service; Host; ServiceGroup; HostGroup | Enable all notifications for services of host x? disable-notifications | - | Service; Host; ServiceGroup; HostGroup | - enable-flap-detection | - | Service; Host; ServiceGroup; HostGroup | - disable-flap-detection | - | Service; Host; ServiceGroup; HostGroup | - enable-event-handler | - | Service; Host | - disable-event-handler | - | Service; Host | -
enable-global-notifications | - | - | - disable-global-notifications | - | - | - enable-global-flap-detection | - | - | - disable-global-flap-detection | - | - | - enable-global-event-handlers | - | - | - disable-global-event-handlers | - | - | - enable-global-performance-data | - | - | - disable-global-performance-data | - | - | - start-global-executing-svc-checks | - | - | - stop-global-executing-svc-checks | - | - | - start-global-executing-host-checks | - | - | - stop-global-executing-host-checks | - | - | - shutdown-process | - | - | - restart-process | - | - | -
Examples:
Reschedule a service check for all services in NOT-OK state:
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/actions/reschedule-check?filter=service.state!=0&type=Service' -X POST | python -m json.tool
{
"results": [
{
"code": 200.0,
"status": "Successfully rescheduled check for icinga.org!http."
},
{
"code": 200.0,
"status": "Successfully rescheduled check for icinga.org!disk."
},
{
"code": 200.0,
"status": "Successfully rescheduled check for icinga.org!disk /."
}
]
}
Event Streams
TODO https://dev.icinga.org/issues/9078
Status and Statistics
Contains a list of sub url endpoints which provide the status and statistics of available and enabled features. Any filters are ignored.
Example for the main url endpoint /v1/status
:
$ curl -k -s -u root:icinga 'https://localhost:5665/v1/status' | python -m json.tool
{
"results": [
{
"name": "ApiListener",
"perfdata": [ ... ],
"status": [ ... ]
},
...
{
"name": "IcingaAplication",
"perfdata": [ ... ],
"status": [ ... ]
},
...
]
}
/v1/status
is always available as virtual status url endpoint.
It provides all feature status information into a collected overview.
Example for the icinga application url endpoint /v1/status/IcingaApplication
:
$ curl -k -s -u root:icinga 'https://localhost:5665/v1/status/IcingaApplication' | python -m json.tool
{
"results": [
{
"perfdata": [],
"status": {
"icingaapplication": {
"app": {
"enable_event_handlers": true,
"enable_flapping": true,
"enable_host_checks": true,
"enable_notifications": true,
"enable_perfdata": true,
"enable_service_checks": true,
"node_name": "icinga.org",
"pid": 59819.0,
"program_start": 1443019345.093372,
"version": "v2.3.0-573-g380a131"
}
}
}
}
]
}
Config Objects
Provides functionality for all configuration object url endpoints provided by config object types:
Url Endpoints | Description |
---|---|
/v1/objects/hosts | Endpoint for retreiving and updating Host objects. |
/v1/objects/services | Endpoint for retreiving and updating Service objects. |
/v1/objects/notifications | Endpoint for retreiving and updating Notification objects. |
/v1/objects/dependencies | Endpoint for retreiving and updating Dependency objects. |
/v1/objects/users | Endpoint for retreiving and updating User objects. |
/v1/objects/checkcommands | Endpoint for retreiving and updating CheckCommand objects. |
/v1/objects/eventcommands | Endpoint for retreiving and updating EventCommand objects. |
/v1/objects/notificationcommands | Endpoint for retreiving and updating NotificationCommand objects. |
/v1/objects/hostgroups | Endpoint for retreiving and updating HostGroup objects. |
/v1/objects/servicegroups | Endpoint for retreiving and updating ServiceGroup objects. |
/v1/objects/usergroups | Endpoint for retreiving and updating UserGroup objects. |
/v1/objects/zones | Endpoint for retreiving and updating Zone objects. |
/v1/objects/endpoints | Endpoint for retreiving and updating Endpoint objects. |
/v1/objects/timeperiods | Endpoint for retreiving and updating TimePeriod objects. |
All object attributes are prefixed with their respective object type.
Example:
host.address
Output listing and url parameters use the same syntax.
API Objects and Joins
Icinga 2 knows about object relations, e.g. when querying a service object the query handler will automatically add the referenced host object and its attributes to the result set. If the object reference is null (e.g. no event_command defined), the joined results not added to the result set.
Note: Select your required attributes beforehand by passing them to your request. The default result set might get huge.
Each joined object will use its own attribute name as prefix for the attribute. There is an exception for multiple objects used in dependencies and zones.
Objects with optional relations (e.g. a host notification does not have services) will not be joined.
Object Type | Object Relations (prefix name) |
---|---|
Service | host, notification, check_command, event_command |
Host | notification, check_command, event_command |
Notification | host, service, command, period |
Dependency | child_host, child_service, parent_host, parent_service, period |
User | period |
Zones | parent |
API Objects and Cluster Config Sync
Newly created or updated objects can be synced throughout your
Icinga 2 cluster. Set the zone
attribute to the zone this object
belongs to and let the API and cluster handle the rest.
If you add a new cluster instance, or boot an instance beeing offline for a while, Icinga 2 takes care of the initial object sync for all objects created by the API.
More information about distributed monitoring, cluster and its configuration can be found here.
List All Objects
Send a GET
request to /v1/objects/hosts
to list all host objects and
their attributes.
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/objects/hosts'
This works in a similar fashion for other config objects.
Create New Config Object
New objects must be created by sending a PUT request. The following parameters need to be passed inside the JSON body:
Parameters | Description |
---|---|
name | Required. Name of the newly created config object. |
templates | Optional. Import existing configuration templates for this object type. |
attrs | Required. Set specific object attributes for this object type. |
If attributes are of the Dictionary type, you can also use the indexer format:
"attrs": { "vars.os": "Linux" }
Example fo creating the new host object google.com
:
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/objects/hosts/google.com' \
-X PUT \
-d '{ "templates": [ "generic-host" ], "attrs": { "address": "8.8.8.8", "check_command": "hostalive", "vars.os" : "Linux" } }' \
| python -m json.tool
{
"results": [
{
"code": 200.0,
"status": "Object was created."
}
]
}
Note: Host objects require the check_command
attribute.
If the configuration validation fails, the new object will not be created and the response body
contains a detailed error message. The following example omits the check_command
attribute required
by the host object.
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/objects/hosts/google.com' \
-X PUT \
-d '{ "attrs": { "address": "8.8.8.8", "vars.os" : "Linux" } }' \
| python -m json.tool
{
"results": [
{
"code": 500.0,
"errors": [
"Error: Validation failed for object 'google.com' of type 'Host'; Attribute 'check_command': Attribute must not be empty."
],
"status": "Object could not be created."
}
]
}
Query Object
Send a GET
request including the object name inside the url.
Example for the host google.com
:
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/objects/hosts/google.com'
You can select specific attributes by adding them as url parameters using ?attrs=...
. Multiple
attributes must be added one by one, e.g. ?attrs=host.address&attrs=host.name
.
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/objects/hosts/google.com?attrs=host.name&attrs=host.address' | python -m json.tool
{
"results": [
{
"attrs": {
"host.address": "8.8.8.8",
"host.name": "google.com"
}
}
]
}
Modify Object
Existing objects must be modifed by sending a POST
request. The following
parameters need to be passed inside the JSON body:
Parameters | Description |
---|---|
name | Optional. If not specified inside the url, this is required. |
templates | Optional. Import existing object configuration templates. |
attrs | Required. Set specific object attributes for this object type. |
If attributes are of the Dictionary type, you can also use the indexer format:
"attrs": { "vars.os": "Linux" }
Example for existing object google.com
:
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/objects/hosts/google.com' \
-X POST \
-d '{ "attrs": { "address": "8.8.4.4", "vars.os" : "Windows" } }' \
| python -m json.tool
{
"results": [
{
"code": 200.0,
"name": "google.com",
"status": "Attributes updated.",
"type": "Host"
}
]
}
Delete Host
You can delete objects created using the API by sending a DELETE
request. Specify the object name inside the url.
Parameters | Description |
---|---|
cascade | Optional. Delete objects depending on the deleted objects (e.g. services on a host). |
Note: Objects created by apply rules (services, notifications, etc) will implicitely require
to pass the cascade
parameter on host object deletion.
Example for deleting the host object google.com
:
$ curl -u root:icinga -k -s 'https://localhost:5665/v1/objects/hosts/google.com?cascade=1' -X DELETE | python -m json.tool
{
"results": [
{
"code": 200.0,
"name": "google.com",
"status": "Object was deleted.",
"type": "Host"
}
]
}
Configuration Management
The main idea behind configuration management is to allow external applications creating configuration packages and stages based on configuration files and directory trees. This replaces any additional SSH connection and whatnot to dump configuration files to Icinga 2 directly. In case you are pushing a new configuration stage to a package, Icinga 2 will validate the configuration asynchronously and populate a status log which can be fetched in a separated request.
Create Config Package
Send a POST
request to a new config package called puppet
in this example. This
will create a new empty configuration package.
$ curl -k -s -u root:icinga -X POST https://localhost:5665/v1/config/packages/puppet | python -m json.tool
{
"results": [
{
"code": 200.0,
"package": "puppet",
"status": "Created package."
}
]
}
Create Configuration to Package Stage
Send a POST
request to the url endpoint /v1/config/stages
including an existing
configuration package, e.g. puppet
.
The request body must contain the files
attribute with the value being
a dictionary of file targets and their content.
The example below will create a new file called test.conf
underneath the conf.d
directory populated by the sent configuration.
The Icinga 2 API returns the package
name this stage was created for, and also
generates a unique name for the package
attribute you'll need for later requests.
Note: This example contains an error (chec_command
), do not blindly copy paste it.
$ curl -k -s -u root:icinga -X POST -d '{ "files": { "conf.d/test.conf": "object Host \"cfg-mgmt\" { chec_command = \"dummy\" }" } }' https://localhost:5665/v1/config/stages/puppet | python -m json.tool
{
"results": [
{
"code": 200.0,
"package": "puppet",
"stage": "nbmif-1441625839-0",
"status": "Created stage."
}
]
}
If the configuration fails, the old active stage will remain active. If everything is successful, the new config stage is activated and live. Older stages will still be available in order to have some sort of revision system in place.
Icinga 2 automatically creates the following files in the main configuration package stage:
File | Description |
---|---|
status | Contains the configuration validation exit code (everything else than 0 indicates an error). |
startup.log | Contains the configuration validation output. |
You can fetch these files via API call after creating a new stage.
List Configuration Packages and their Stages
List all config packages, their active stage and other stages. That way you may iterate of all of them programmatically for older revisions and their requests.
The following example contains one configuration package puppet
.
The latter already has a stage created, but it is not active.
$ curl -k -s -u root:icinga https://localhost:5665/v1/config/packages | python -m json.tool
{
"results": [
{
"active-stage": "",
"name": "puppet",
"stages": [
"nbmif-1441625839-0"
]
}
]
}
List Configuration Packages and their Stages
Sent a GET
request to the url endpoint /v1/config/stages
including the package
(puppet
) and stage (nbmif-1441625839-0
) name.
$ curl -k -s -u root:icinga https://localhost:5665/v1/config/stages/puppet/nbmif-1441625839-0 | python -m json.tool
{
"results": [
...
{
"name": "startup.log",
"type": "file"
},
{
"name": "status",
"type": "file"
},
{
"name": "conf.d",
"type": "directory"
},
{
"name": "zones.d",
"type": "directory"
},
{
"name": "conf.d/test.conf",
"type": "file"
}
]
}
Fetch Configuration Package Stage Files
Send a GET
request to the url endpoint /v1/config/files
including
the package name, the stage name and the relative path to the file.
Note: You cannot use dots in paths.
You can fetch a list of existing files in a configuration stage and then specifically request their content.
The following example fetches the faulty configuration inside conf.d/test.conf
for further analysis.
$ curl -k -s -u root:icinga https://localhost:5665/v1/config/files/puppet/nbmif-1441625839-0/conf.d/test.conf
object Host "cfg-mgmt" { chec_command = "dummy" }
Note: The returned files are plain-text instead of JSON-encoded.
Configuration Package Stage Errors
Now that we don’t have an active stage for puppet
yet seen here,
there must have been an error.
Fetch the startup.log
file and check the config validation errors:
$ curl -k -s -u root:icinga https://localhost:5665/v1/config/files/puppet/imagine-1441133065-1/startup.log
...
critical/config: Error: Attribute 'chec_command' does not exist.
Location:
/var/lib/icinga2/api/packages/puppet/imagine-1441133065-1/conf.d/test.conf(1): object Host "cfg-mgmt" { chec_command = "dummy" }
^^^^^^^^^^^^^^^^^^^^^^
critical/config: 1 error
The output is similar to the manual configuration validation.