Skip to content

[pull] master from DataDog:master - #651

Merged
pull[bot] merged 3 commits into
ConnectionMaster:masterfrom
DataDog:master
Jul 10, 2026
Merged

[pull] master from DataDog:master#651
pull[bot] merged 3 commits into
ConnectionMaster:masterfrom
DataDog:master

Conversation

@pull

@pull pull Bot commented Jul 10, 2026

Copy link
Copy Markdown

See Commits and Changes for more details.


Created by pull[bot] (v2.0.0-alpha.4)

Can you help keep this open source service alive? 💖 Please sponsor : )

piochelepiotr and others added 3 commits July 10, 2026 03:22
…24471)

* kafka_consumer: store broker_timestamps as marshal instead of json

The DSM broker_timestamps cache was serialized with json.dumps on every check
run. At high partition counts json.dumps is the dominant allocation source
(profiler flagged _save_broker_timestamps -> json.dumps -> iterencode as the top
allocator, feeding glibc arena fragmentation and CCR OOMs, #incident-57455).

Persist with marshal (base64-wrapped to ride the string persistent-cache API)
instead. marshal is a compact binary codec that avoids JSON's per-number text
encoding, cutting serialization ~100x on the save path with no change to emitted
metrics. The on-disk format changes and is version-specific, so a cache written
by an older agent is discarded once (rebuilt next run).

Tests seed and read the cache in the marshal format with integer offset keys.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Add changelog entry

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(incident-io): add account_config.json with API key auth

* Trigger CODEOWNERS re-evaluation

* Update label for API key in account_config.json

* Update account_config.json
* cockroachdb: fix stale metric names in metadata.csv

METRIC_MAP renames the six changefeed_*_hist_nanos raw metric families by
dropping the _nanos suffix (e.g. changefeed.checkpoint_hist), but
metadata.csv still listed the old _nanos-suffixed names for the
bucket/count/sum rows, and had no entries for go.scheduler_latency. Add a
test that runs without histogram_buckets_as_distributions, which is what
surfaces the mismatch since the default instance bypasses classic
bucket/count/sum submission entirely.

Environment: Datadog workspace

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

* Add changelog entry for #24461

Environment: Datadog workspace

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

* Remove unneeded changelog entry for #24461

metadata.csv is not shipped with the Agent, so no changelog entry is required for this change.

Environment: Datadog workspace

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
@pull pull Bot locked and limited conversation to collaborators Jul 10, 2026
@pull pull Bot added the ⤵️ pull label Jul 10, 2026
@pull
pull Bot merged commit 38ce43f into ConnectionMaster:master Jul 10, 2026
1 check passed
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants