Configuration
Configuration File
Configuration is contained in config.yaml, either in the same directory as application or in /config/config.yaml.
While file needs to be a valid yaml file, quite permissive parsing is used and thus attempt will be made to set non-parsable values to a sane default and/or unrecognized elements will be ignored.
Example of config file is as follows:
checks:
- title: Always success
kind: dummy
- title: Random
name: random
kind: random
- title: AASeq ping
kind: ping
target: aaseq.com
timeout: 0.1
- title: Random Health Probe
kind: get
target: http://localhost:8089/healthz/random
success: 3
failure: 2
period: 20
timeout: 0.1
web:
title: Status
refresh: 5
logging:
console: verbose
file:
level: warning
path: /var/log/revidere.log
seq:
level: information
url: http://localhost:5341
checks
This is a node containing a sequence. Each sequence member defines one health check.
Common
Each member of sequence can contain the following keys: * kind: defines kind of health check; must be specified * target: defines target of a check; the exact syntax depends on the check kind * name: controls the name of check when using healthz path; each name must be unique and consists only of letter, number, dash (-), and underscore (_) characters * title: this key controls the title of the given check for the web interface; if value is not present it will default to either check name (if available) or check kind * period: interval between checks; must be between 1 second and 10 minutes; default is 10 seconds * timeout: timeout for each check; it must be between 10 milliseconds and 10 seconds; its value cannot be larger than period; default is 5 seconds * success: number of consecutive successful attempts before a target is considered healthy; must be between 1-10 seconds; default is 2 * failure: number of consecutive failed attempts before a healthy target is considered healthy; must be between 1-10 seconds; default is 2 * visible: if set to false, check will not be visible in GUI; default is true * break: if set to true, a new line will start after it
This node is mandatory.
kind: get/head/post/put/delete
Key target is a mandatory and its value must be URL that request is sent to. Kind matches the HTTP method to be used.
Example:
- kind: get
target: http://example.com
- kind: head
target: http://example.net
kind: ping
Key target is a mandatory and its value must be either hostname or ip that ping request will be sent to.
Example:
- kind: ping
target: 1.1.1.1
kind: tcp
Key target is a mandatory and its value must be either hostname or ip that followed by colon and port number.
Example:
- kind: tcp
target: 1.1.1.1:443
kind: dummy
This kind is used for troubleshooting. It always returns true.
Example:
- kind: dummy
kind: random
This kind is used for troubleshooting. It returns random health check outcomes.
The same target will result in the same random number seed. If target is not specified, the current time will be used.
Example:
- kind: random
web
This node defines mappings for built-in web server. Its usage is optional.
Key prefix controls location where web server will be listening for requests. By default this will be on port 8089 (http://*:8089/). If multiple prefixes are required, they can be specified as an sequence or space-separated string. This key is optional.
Key title controls which title will webpage have (if custom title is requested). This key is optional.
Key refresh controls how often page will refresh in a browser. Please note that this only impacts web interface and not health checks. Default value is 10 seconds. This key is optional.
Example:
web:
prefix: http://*:8089
title: Status
refresh: 10
logging
This is a mapping node defining logging setup.
This node is optional.
logging (Common)
Each logging subnode has level key in common.
Key level can contain the following values: * none: no logging will be performed * verbose: full logging is enabled * debug: debug logging is enabled * information: informational messages will be shown * warning: warning messages not impacting system health * error: issues having a significant impact to the application * panic: unrecoverable application errors
Note that each level implicitly also includes higher severity levels, e.g. using warning will print warning, error, and panic.
logging/console
Mappings in this node define logging properties for console output. If there is no console node, application will default to using information console level logging.
Example:
logging:
console:
level: verbose
For console node only, this can be shortened:
logging:
console: verbose
logging/file
Mappings in this node define logging properties for file output. If not specified, default log level of debug is used.
Key path must contain the name of the log file.
Key interval defines rolling interval for the file and can be one of the following values: * infinite: no file rolling will occur * year: file will rollover once a year * month: file will rollover once a month * day: file will rollover once a day (default) * hour: file will rollover once a hour * minute: file will rollover once a minute
Key retain controls how many files will be retained after rollover. Default is 7.
Key buffered controls if buffering will be used. Default is true.
Example:
logging:
file:
level: warning
path: /var/log/revidere.log
interval: hour
retain: 24
logging/seq
Mappings in this node define logging properties for seq output. If not specified, default log level of info is used.
Key url must contain the server accepting the log messages.
Example:
logging:
seq:
level: warning
url: http://localhost:5341
Environment Variable
Limited configuration can be done using environment variables. This is not intended for long-term usage.
CHECKS
Health checks can be specified in comma-separated list. Each health check target will use its kind as a prefix and its target as host/path parameters. No custom configuration will be available and for HTTP checks, only GET method is supported.
Example:
CHECKS=ping://8.8.4.4,http://aaseq.com