Prometheus v2.30.0 Release Notes

Release Date: 2021-09-14 // over 2 years ago
    • [FEATURE] experimental TSDB: Snapshot in-memory chunks on shutdown for faster restarts. Behind --enable-feature=memory-snapshot-on-shutdown flag. #7229
    • [FEATURE] experimental Scrape: Configure scrape interval and scrape timeout via relabeling using __scrape_interval__ and __scrape_timeout__ labels respectively. #8911
    • โฑ [FEATURE] Scrape: Add scrape_timeout_seconds and scrape_sample_limit metric. Behind --enable-feature=extra-scrape-metrics flag to avoid additional cardinality by default. #9247 #9295
    • [ENHANCEMENT] Scrape: Add --scrape.timestamp-tolerance flag to adjust scrape timestamp tolerance when enabled via --scrape.adjust-timestamps. #9283
    • [ENHANCEMENT] Remote Write: Improve throughput when sending exemplars. #8921
    • [ENHANCEMENT] TSDB: Optimise WAL loading by removing extra map and caching min-time #9160
    • [ENHANCEMENT] promtool: Speed up checking for duplicate rules. #9262/#9306
    • ๐Ÿ“œ [ENHANCEMENT] Scrape: Reduce allocations when parsing the metrics. #9299
    • ๐Ÿณ [ENHANCEMENT] docker_sd: Support host network mode #9125
    • ๐Ÿ›  [BUGFIX] Exemplars: Fix panic when resizing exemplar storage from 0 to a non-zero size. #9286
    • [BUGFIX] TSDB: Correctly decrement prometheus_tsdb_head_active_appenders when the append has no samples. #9230
    • ๐Ÿ›  [BUGFIX] promtool rules backfill: Return 1 if backfill was unsuccessful. #9303
    • ๐Ÿ›  [BUGFIX] promtool rules backfill: Avoid creation of overlapping blocks. #9324
    • ๐Ÿ›  [BUGFIX] config: Fix a panic when reloading configuration with a null relabel action. #9224