Add geoip database metrics to /node/stats API

[kaisecheng]

Release notes

Add geoip database metrics to /node/stats

What does this PR do?

This PR exposes geoip database metrics to Logstash localhost:9600/node/stats API

Sample output

"geoip" : {
    "database" : {
       “ASN”: {
          "status" : "healthy",
          "fail_check_in_days" : 0,
          "download_at" : "2021-06-21T16:06:54+02:00"
       },
       “City”: {
          "status" : "healthy",
          "fail_check_in_days" : 0,
          "download_at" : "2021-06-21T16:06:54+02:00"
       }
    },
    "download" : {
        "successes" : 15,
        "failures" : 1,
        "last_check_at" : "2021-06-21T16:07:03+02:00",
        "status" : "succeeded"
    }
}

Schema

geoip
  database
    ASN/CIty
      status:
        - init, initial CC database status
        - healthy, using up-to-date EULA database
        - to_be_expired, 25 days without calling service
        - expired, 30 days without calling service
      fail_check_in_days: nr. of days LS fail to call service since the last success
      download_at: the timestamp of the last database download
  download
    successes: the number of successful checks and downloads. When all databases are updated or pass the check, the counter +1
    failures: the number of failed attempts. When one of the databases fails to check or update, the counter +1
    last_check_at: the timestamp of last check
    status:
      - checking, check and download at the moment
      - succeeded, last download succeed
      - failed, last download failed

Why is it important/What is the impact to the user?

It allows system administrators to monitor database status through metric API instead of scanning the log.
When they see database status to_be_expired, they should allow Logstash to access the internet to download the latest database to prevent the expired status that fails the plugin

Checklist

• My code follows the style guidelines of this project
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have made corresponding change to the default configuration files (and/or docker env variables)
• I have added tests that prove my fix is effective or that my feature works

Author's Checklist

• [ ]

How to test this PR locally

please check the geoip session of curl -XGET 'localhost:9600/_node/stats'

Related issues

Use cases

Screenshots

Logs

[andsel]

LGTM

Left just a couple of notes on possible improvements for readability

[andsel]

I think this assignment could be extracted in a method update_download_status(:checking). In this way the intention is more evident without the read immediately deep into the knowledge of metrics.

[andsel]

Here we could also use update_download_status so in the flow of download a DB becomes more evident the status change points.

[andsel]

Maybe this could be split in two methods, one for the counter and the other for the status, so that the status update points becomes evident in the download flow

[roaksoax]

After discussion, we decided to improve the wording to reflect the following:

geoip -> geoip_download_manager (7.14)
database status:

• init -> remains for the time being
• cc - add a new state that reflects that we are using the bundled db because it never was able to download from the service.
• healthy -> up-to-date (7.14)
• download_at -> last_updated (7.14)

download -> download_stats ? (7.14)

• last_failed_at -> adds a date with the time it last failed.
• last_error_msg -> adds the last error message, if any.
• status: checking -> updating (for the future we can consider checking, downloading, replacing/updating). (7.14)

[karenzone]

Docs added: #13019