Read the status output of the nginx_upstream_check (https://github.com/yaoweibin/nginx_upstream_check_module).
This module can periodically check the servers in the Nginx's upstream with configured request and interval to determine
if the server is still available. If checks are failed the server is marked as "down" and will not receive any requests
until the check will pass and a server will be marked as "up" again.
The status page displays the current status of all upstreams and servers as well as number of the failed and successful
checks. This information can be exported in JSON format and parsed by this input.
Configuration
## An URL where Nginx Upstream check module is enabled
## It should be set to return a JSON formatted response
url = "http://127.0.0.1/status?format=json"
## HTTP method
# method = "GET"
## Optional HTTP headers
# headers = {"X-Special-Header" = "Special-Value"}
## Override HTTP "Host" header
# host_header = "check.example.com"
## Timeout for HTTP requests
timeout = "5s"
## Optional HTTP Basic Auth credentials
# username = "username"
# password = "pa$$word"
## Optional TLS Config
# tls_ca = "/etc/telegraf/ca.pem"
# tls_cert = "/etc/telegraf/cert.pem"
# tls_key = "/etc/telegraf/key.pem"
## Use TLS but skip chain & host verification
# insecure_skip_verify = false
Measurements & Fields
- Measurement
- fall (The number of failed server check attempts, counter)
- rise (The number of successful server check attempts, counter)
- status (The reporter server status as a string)
- status_code (The server status code. 1 - up, 2 - down, 0 - other)
The "status_code" field most likely will be the most useful one because it allows you to determine the current
state of every server and, possible, add some monitoring to watch over it. InfluxDB can use string values and the
"status" field can be used instead, but for most other monitoring solutions the integer code will be appropriate.
- All measurements have the following tags:
- name (The hostname or IP of the upstream server)
- port (The alternative check port, 0 if the default one is used)
- type (The check type, http/tcp)
- upstream (The name of the upstream block in the Nginx configuration)
- url (The status url used by telegraf)
Example Output
When run with:
./telegraf --config telegraf.conf --input-filter nginx_upstream_check --test
It produces:
* Plugin: nginx_upstream_check, Collection 1
> nginx_upstream_check,host=node1,name=192.168.0.1:8080,port=0,type=http,upstream=my_backends,url=http://127.0.0.1:80/status?format\=json fall=0i,rise=100i,status="up",status_code=1i 1529088524000000000
> nginx_upstream_check,host=node2,name=192.168.0.2:8080,port=0,type=http,upstream=my_backends,url=http://127.0.0.1:80/status?format\=json fall=100i,rise=0i,status="down",status_code=2i 1529088524000000000