search

pelias / docker

Run the Pelias geocoder in docker containers, including example projects.
MIT License
314 stars 217 forks source link

Self hosted planet build result deviation #331

Closed MALKGITHUBER closed 7 months ago

MALKGITHUBER commented 7 months ago

Hi Team, First of all i want to appreaciate and Thanks all of the pelias Team efforts in making the pelias component such a great one,

Here i got some result deviation while building planet data

screenshot-12

pelias-dashboard build

newpeliasdashboard

There is some deviation in

  1. doc store size and,

  2. address Layer according to the pelias-dashoard,

    ran the build multiple times, getting the same result

is there anything am i missing!!

Thanks in advance

missinglink commented 7 months ago

Hi @MALKGITHUBER, the data available via the APIs provided by https://geocode.earth contain some additions which aren't available in the free data we provide.

Your installation contains 689.1 million addresses which should be more than sufficient for most workloads.

MALKGITHUBER commented 7 months ago

Hi @MALKGITHUBER, the data available via the APIs provided by https://geocode.earth contain some additions which aren't available in the free data we provide.

Your installation contains 689.1 million addresses which should be more than sufficient for most workloads.

Thanks for the response

doc store size might not be problem right?

missinglink commented 7 months ago

doc store size might not be problem right

lower is better

MALKGITHUBER commented 7 months ago

size

Thanks for your efforts in clarifying the doubt

MALKGITHUBER commented 6 months ago

Hi missinglink , How much acceptance test success rate is good for self hosted planet build ?

Here is my result, is this good to go

Aggregate test results Pass: 355 Improvements: 9 Expected Failures: 116 Placeholders: 0 Regressions: 356 Total tests: 836 Took 37688ms Test success rate 57.42%

FATAL ERROR: 356 regression(s) detected.