Stackdriver Exporter Versions Save

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

• [BUGFIX] Fix default listening port #229

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

• [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

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

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

• [FEATURE] Autodiscover Google Poject ID #62

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

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

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