Skip to content

Latest commit

 

History

History
123 lines (87 loc) · 6.15 KB

README.md

File metadata and controls

123 lines (87 loc) · 6.15 KB
Raw

gaarf exporter

Prometheus exporter for Google Ads metrics with customizable metrics collectors.

Installation and usage

Locally

  1. Install gaarf-exporter from pip:
pip install gaarf-exporter
  1. Run gaarf-exporter:
gaarf-exporter

Docker

docker run --network=host \
  -v /path/to/google-ads.yaml:/google-ads.yaml \
  -v `pwd`/gaarf_exporter.yaml:/app/gaarf_exporter.yaml \
  gaarf_exporter

docker run --network=host gaarf_exporter \
  --config gs://path/to/gaarf_config.yaml \
  --ads-config gs://path/to/google-ads.yaml

By default it will start http_server on localhost:8000 and will push some basic metrics to it.

Customization

  • --ads-config - path to google-ads.yaml

    ads-config can be taken from local storage or remote storage (gs, s3, azure, ssh, stfp, scrp, hdfs, webhdfs).

  • --config - path to gaarf_exporter.yaml

    config can be taken from local storage or remote storage (same as --ads-config).

  • --collectors - names of one or more collectors (separated by comma).
  • --http_server.address - address of your http server (localhost by default)
  • --http_server.port - port of your http server (8000 by default)
  • --pushgateway.address - address of your pushgateway service (None by default)
  • --pushgateway.port - port of your pushgateway (None by default)
  • --delay-minutes - delay in minutes between scrapings (15 by default)

Customizing with macros:

Customizing fetching dates

By default gaarf-exporter fetches performance data for TODAY; if you want to customize it you can provide optional flags:

  • --macro.start_date=:YYYYMMDD-N, where N is number of days starting from today
  • --macro.end_date=:YYYYMMDD-M, where N is number of days starting from today

It will add an additional metric to be exposed to Prometheus *_n_days (i.e. googleads_clicks_n_days).

Collectors

You can specify collectors with --collectors <collector_name> CLI argument. Some collectors available by default, other you need to specify explicitly.
There are two types of collectors - registry (contains other collectors grouped logically) and collectors themselves.

default registry.

  • performance - extracts clicks, impressions, cost, conversions on by ad_network and ad_group_id
  • conversion_action - extracts all_conversions by conversion_id and account_id
  • mapping - performance mapping between ad_group_id, ad_group_name, campaign_id, campaign_name, campaign_status, account_id, account_name

disapprovals registry.

  • ad_disapprovals - extracts approval_status, review_status, topic, topic_type by ad_group_id and ad_id only for not approved ads
  • ad_group_ad_asset_disapprovals - extracts approval_status, review_status, topic, topic_type by ad_group_id and asset_id only for enabled assets.
  • sitelink_disapprovals - extracts approval_status, review_status, topic, topic_type by asset_id, sitelink_title and both sitelink descriptions only for not approved sitelinks.
  • pmax_disapprovals - extracts approval_status, review_status, topic, topic_type by asset_id, asset_group_id for active asset group assets.

app registry.

  • app_campaign_mapping - performs mapping between campaign_id, app_id, app_store, and bidding_strategy only for active campaigns.
  • app_asset_mapping - performs mapping between asset_id and its type, source, and content (name, text, video_id) only for app assets (HTML5, TEXT, IMAGE, VIDEO).
  • asset_performance - extracts clicks, impressions, cost, installs, inapps, and conversions_value by ad_group_id, ad_network and asset_id
  • asset_perf_label - extracts performance_label by ad_group_id and asset_id

pmax registry.

  • pmax_mapping - performs mapping between asset_group_id, asset_group_name, and meta information on campaign and account only for active campaigns and enabled asset groups.
  • pmax_performance - extracts clicks, impressions, cost, installs, inapps, and conversions_value by asset_group_id.
  • pmax_disapprovals - extracts approval_status, review_status, topic, topic_type by asset_id, asset_group_id for active asset group assets.

search registry.

  • search_terms - extracts clicks, impressions, cost, conversions on by search_term and ad_group_id
  • search_terms_conversion_split - extracts all_conversions by search_term and conversion_id on ad_group_id level
  • keywords - extracts clicks, impressions, cost, conversions, and historical auality_score by keyword and match_type on ad_group level.
  • keywords_conversion_split - extracts all_conversions by keyword and match_type on ad_group level.

placements registry.

  • placements - extracts clicks, impressions, cost, conversions on by placement_name and placement_type for each account.
  • placements_conversion_split - extracts all_conversions by placement_type and placement_type for each account.

demographics registry.

  • age - extracts clicks, impressions, cost, conversions by age_range and campaign_id
  • age_conversion_split - extracts all_conversions by age_range and conversion_id on campaign_id level
  • gender - extracts clicks, impressions, cost, conversions by gender and campaign_id
  • gender_conversion_split - extracts all_conversions by gender and conversion_id on campaign_id level

geo registry.

  • user_location - extracts clicks, impressions, cost, conversions by country_id and campaign_id only for active campaigns.
  • user_location_conversion_split - extracts all_conversions by country_id and campaign_idonly for active campaigns.

collectors without registry.

  • bid_budgets - extracts current values of bid (target_cpa, target_roas) and campaign budgets.
  • bids - extracts current values of bid (target_cpa, target_roas).
  • budgets - extracts current values of campaign budgets.
  • account_status - extracts customer_status for each account.
  • campaign_service_status - extracts primary_status for each campaign.