Skip to content

Latest commit

 

History

History
93 lines (59 loc) · 3.81 KB

CHANGELOG.md

File metadata and controls

93 lines (59 loc) · 3.81 KB
Raw

master / unreleased

0.17.0 / 2024-11-04

Deprecation notice: The comma delimited flags google.project-id and monitoring.metrics-type-prefixes are being replaced by repeatable flags google.project-ids and monitoring.metrics-prefixes. The comma delimited flags will be supported for at least one more release.

  • [CHANGE] Migrate logging to promslog #378
  • [ENHANCEMENT] Sanitize metric type prefixes to prevent duplicate metrics #319
  • [ENHANCEMENT] Add project ID to all logs from the collector #362
  • [FEATURE] Add support for specifying comma-delimited string flags as repeatable flags #355

0.16.0 / 2024-07-15

  • [FEATURE] Add ErrorLogger for promhttp #277
  • [ENHANCEMENT] Add more info about filters to docs and rename struct fields #198

0.15.1 / 2024-05-15

  • [BUGFIX] Fix histogram merge #324

0.15.0 / 2024-03-07

  • [FEATURE] Add projects query #243
  • [ENHANCEMENT] Refactor delta logic for library usage #190

0.14.1 / 2023-05-26

  • [BUGFIX] Fix default listening port #229

0.14.0 / 2023-05-26

  • [FEATURE] cache descriptors to reduce API calls #218

0.13.0 / 2023-01-25

  • [FEATURE] Add monitoring.aggregate-deltas and monitoring.aggregate-deltas-ttl flags which allow aggregating DELTA metrics as counters instead of a gauge #168
  • [FEATURE] Add web.stackdriver-telemetry-path flag. When configured the stackdriver metrics go to this endpoint and web.telemetry-path contain just the runtime metrics. #173
  • [ENHANCEMENT] Make Stackdriver main collector more library-friendly #157
  • [BUGFIX] Fixes suspected duplicate label panic for some GCP metric #153
  • [BUGFIX] Metrics-ingest-delay bugfix #151
  • [BUGFIX] Fix data race on metricDescriptorsFunction start and end times #158

0.12.0 / 2022-02-08

Breaking Changes:

The exporter nolonger supports configuration via ENV vars. This was a non-standard feature that is not part of the Prometheus ecossystem. All configuration is now handled by the existing command line arguments.

  • [CHANGE] Cleanup non-standard ENV var setup #142
  • [FEATURE] Add support to include ingest delay when pull metrics #129
  • [FEATURE] Add monitoring.filters flag #133
  • [ENHANCEMENT] Setup exporter metrics only once when we can #124

0.11.0 / 2020-09-02

  • [CHANGE] Do not treat failure to collect metrics as fatal #102
  • [FEATURE] Add support for multiple google project IDs #105

0.10.0 / 2020-06-28

  • [FEATURE] Autodiscover Google Poject ID #62

0.9.1 / 2020-06-02

  • [BUGFIX] Fix report time missing for histogram metrics #94

0.9.0 / 2020-05-26

  • [CHANGE] Add stackdriver timestamp to metrics #84
  • [CHANGE] Fix collect param name #91

0.8.0 / 2020-05-13

  • [CHANGE] Treat failure to collect metric as fatal #83
  • [CHANGE] Switch logging to promlog #88
  • [FEATURE] Add metrics prefix collect URL param #87

0.7.0 / 2020-05-01

  • [CHANGE] Remove deprecated monitoring.New() use. #76
  • [ENHANCEMENT] Server-side selection of project's metrics #53
  • [BUGFIX] Ensure metrics are fetched once for each metric descriptor #50

0.6.0 / 2018-12-02

Google Stackdriver Prometheus Exporter v0.6.0:

  • Added a collector.fill-missing-labels flag to fill missing metrics labels with empty strings in order to avoid label dimensions inconsistent failure (PR #23)
  • Added stackdriver.max-retries, stackdriver.http-timeout, stackdriver.max-backoff, stackdriver.backoff-jitter, andstackdriver.retry-statuses flags to allow exponential backoff and retries on stackdriver api (PR #35)
  • Added a monitoring.drop-delegated-projects flag which allows one to disable metrics collection from delegated projects (PR #40)
  • Fix segmentation fault on missing credentials (PR #42)