Skip to content

Latest commit

 

History

History
181 lines (150 loc) · 9.18 KB

README.md

File metadata and controls

181 lines (150 loc) · 9.18 KB
Raw

Circonus Agent

The etc directory is used for the main configuration of the Circonus agent, as well as, configuration files for builtin collectors.

Main configuration

File name: circonus-agent.(json|toml|yaml)

An example configuration, with default values, can be retrieved using the --show-config=(json|toml|yaml).

Configuration file quick start

Run one of the following (from the base directory where the agent was installed) and edit the resulting configuration file:

sbin/circonus-agentd --show-config=json > etc/circonus-agent.json.tmp
sbin/circonus-agentd --show-config=toml > etc/circonus-agent.toml.tmp
sbin/circonus-agentd --show-config=yaml > etc/circonus-agent.yaml.tmp

or, on Windows:

sbin\circonus-agentd.exe --show-config=json > etc\circonus-agent.json.tmp
sbin\circonus-agentd.exe --show-config=toml > etc\circonus-agent.toml.tmp
sbin\circonus-agentd.exe --show-config=yaml > etc\circonus-agent.yaml.tmp

Edit the resulting file to customize configuration settings. When done, rename file to remove the .tmp extension. (e.g. mv etc/circonus-agent.json.tmp etc/circonus-agent.json)

Collector configurations

Three formats are supported (json, toml, yaml) for collector configurations. Collector configurations should be stored in the etc directory where the agent was installed (for example: /opt/circonus/agent/etc or C:\circonus-agent\etc).

Linux

ProcFS collectors

All ProcFS collectors have a basic set of configuration options:

Option Type Default Description
id string name of collector ID/Name of the collector (used as prefix for metrics).
metrics_enabled array of strings empty list of metrics which are enabled (to be collected)
metrics_disabled array of strings empty list of metrics which are disabled (should NOT be collected)
metrics_default_status string enabled how a metric NOT in the enabled/disabled lists should be handled ("enabled" or "disabled")
run_ttl string empty indicating collector will run no more frequently than TTL (e.g. "10s", "5m", etc. - for expensive collectors)

Additionally, each collector may have more configuration options specific to what is being collected. (e.g. include/exclude regular expression for items such as network interfaces, disks, etc.)

  • CPU
    • ID: cpu
    • Config file: cpu_collector.(json|toml|yaml)
    • Options:
      • report_all_cpus string, include all cpus, not just total (default "false")
  • Disk stats
    • ID: diskstats
    • Config file: diskstats_collector.(json|toml|yaml)
    • Options:
      • include_regex string, regular expression for disk inclusion - default .+
      • exclude_regex string, regular expression for disk exclusion - default empty
  • Network interfaces
    • ID: if
    • Config file: if_collector.(json|toml|yaml)
    • Options:
      • include_regex string, regular expression for interface inclusion - default .+
      • exclude_regex string, regular expression for interface exclusion - default lo
  • Memory
    • ID: vm
    • Config file: vm_collector.(json|toml|yaml)
    • Options: only the common options
  • System load
    • ID: loadavg
    • Config file: loadavg_collector.(json|toml|yaml)
    • Options: only the common options

Windows

WMI

All WMI collectors have a basic set of configuration options:

Option Type Default Description
id string name of collector ID/Name of the collector (used as prefix for metrics).
metrics_enabled array of strings empty list of metrics which are enabled (to be collected)
metrics_disabled array of strings empty list of metrics which are disabled (should NOT be collected)
metrics_default_status string enabled how a metric NOT in the enabled/disabled lists should be handled ("enabled" or "disabled")
metric_name_regex string [^a-zA-Z0-9.-_:] regular expression of valid characters for the metric names
metric_name_char string _ used for replacing invalid characters in a metric name (those not matching metric_name_regex)
run_ttl string empty indicating collector will run no more frequently than TTL (e.g. "10s", "5m", etc. - for expensive collectors)

Additionally, each collector may have more configuration options specific to what is being collected. (e.g. include/exclude regular expression for items such as network interfaces, disks, processes, file systems, etc.)

  • Cache
    • ID: cache
    • Config file: cache_collector.(json|toml|yaml)
    • Options: only the common options
  • Disk
    • ID: disk
    • Config file: disk_collector.(json|toml|yaml)
    • Options:
      • logical_disks string(true|false), include logical disks (default "true")
      • physical_disks string(true|false), include physical disks (default "true")
      • include_regex string, regular expression for disk inclusion - default .+
      • exclude_regex string, regular expression for disk exclusion - default empty
  • Memory
    • ID: memory
    • Config file: memory_collector.(json|toml|yaml)
    • Options: only the common options
  • Network interfaces
    • ID: interface
    • Config file: interface_collector.(json|toml|yaml)
    • Options:
      • include_regex string, regular expression for interface inclusion - default .+
      • exclude_regex string, regular expression for interface exclusion - default empty
  • IP network protocol
    • ID: ip
    • Config file: ip_collector.(json|toml|yaml)
    • Options:
      • enable_ipv4 string(true|false), include IPv4 metrics - default "true"
      • enable_ipv6 string(true|false), include IPv6 metrics - default "true"
  • TCP network protocol
    • ID: tcp
    • Config file: tcp_collector.(json|toml|yaml)
    • Options:
      • enable_ipv4 string(true|false), include IPv4 metrics - default "true"
      • enable_ipv6 string(true|false), include IPv6 metrics - default "true"
  • UDP network protocol
    • ID: udp
    • Config file: udp_collector.(json|toml|yaml)
    • Options:
      • enable_ipv4 string(true|false), include IPv4 metrics - default "true"
      • enable_ipv6 string(true|false), include IPv6 metrics - default "true"
  • Objects
    • ID: objects
    • Config file: objects_collector.(json|toml|yaml)
    • Options: only the common options
  • Paging file
    • ID: paging_file
    • Config file: paging_file_collector.(json|toml|yaml)
    • Options:
      • include_regex string, regular expression for file inclusion - default .+
      • exclude_regex string, regular expression for file exclusion - default empty
  • Processors
    • ID: processor
    • Config file: processor_collector.(json|toml|yaml)
    • Options:
      • report_all_cpus string, include all cpus, not just total (default "true")
  • Processes
    • ID: processes
    • NOTE: disabled by default (28 metrics per process)
    • Config file: processes_collector.(json|toml|yaml)
    • Options:
      • include_regex string, regular expression for process inclusion - default .+
      • exclude_regex string, regular expression for process exclusion - default empty

Common

Prometheus collector

Collect from endpoints exposing prometheus formatted metrics. The Prometheus collector is enabled by default. It is automatically disabled if no configuration file is found.

ID: prometheus Config file: prometheus_collector.(json|toml|yaml) Options:

Option Type Default Description
metrics_enabled array of strings empty list of metrics which are enabled (to be collected)
metrics_disabled array of strings empty list of metrics which are disabled (should NOT be collected)
metrics_default_status string enabled how a metric NOT in the enabled/disabled lists should be handled ("enabled" or "disabled")
run_ttl string empty indicating collector will run no more frequently than TTL (e.g. "10s", "5m", etc. - for expensive collectors)
urls array of urldefs empty required, without any URLs the collector is disabled
URL definition (urldefs)
id string empty required, used as prefix for metrics from this URL
url string url required, URL which responds with Prometheus text format metrics
ttl string 30s optional, timeout for the request