Possible approach to solve GeoIP update?

["Yerex, Tom" <to...@ubc.ca>]

Good afternoon,

Reviewing hxxps://issues.apache.org/jira/projects/METRON/issues/METRON-2340 I'm attempting to sketch out a rough solution and I would like guidance from more experienced minds.

Maxmind releases code that allows you to build your own mmdb database. The tests I can see in Metron, besides the download, seem to try a few locations like Milton and London. Would it be sufficient to generate geoip databases with those (limited), test locations or alternatively generate geoip databases with hypothetical locations and then host those locations in git for the purpose of the download and testing?

The documentation for the settings in Ambari for Metron GeoIP updates could then be slightly updated to add that someone needs to get their own key for GeoIP updates?

Cheers,

Tom.

Re: Possible approach to solve GeoIP update?

[Nick Allen <ni...@nickallen.org>]

That seems like a viable solution to me.



On Thu, Apr 9, 2020 at 7:58 PM Yerex, Tom <to...@ubc.ca> wrote:

> Good afternoon,
>
> Reviewing hxxps://
> issues.apache.org/jira/projects/METRON/issues/METRON-2340 I'm attempting
> to sketch out a rough solution and I would like guidance from more
> experienced minds.
>
> Maxmind releases code that allows you to build your own mmdb database. The
> tests I can see in Metron, besides the download, seem to try a few
> locations like Milton and London. Would it be sufficient to generate geoip
> databases with those (limited), test locations or alternatively generate
> geoip databases with hypothetical locations and then host those locations
> in git for the purpose of the download and testing?
>
> The documentation for the settings in Ambari for Metron GeoIP updates
> could then be slightly updated to add that someone needs to get their own
> key for GeoIP updates?
>
> Cheers,
>
> Tom.
>