[AGNTLOG-392] Logs Agent HTTP Retry with Exponential Backoff

[ellataira]

What does this PR do?

Implements automatic HTTP transport upgrade under an exponential backoff for the Logs Agent when HTTP connectivity is restored after an initial fallback to TCP, along with improved telemetry.

Smart HTTP Retry and Upgrade Mechanism:

• When the Logs Agent fails initial HTTP connectivity checks, it falls back to TCP protocol
• If configuration permits (no force_use_tcp, socks5_proxy_address, or additional_endpoints set), the agent periodically retries HTTP connectivity
• Once HTTP connectivity is verified, the agent automatically upgrades to HTTP transport
• Commit to HTTP: After successful HTTP connectivity check, the agent commits to HTTP and continuously retries HTTP transport even if restart attempts fail, ensuring eventual upgrade
• Configurable logs_config.http_connectivity_retry_interval_max for exponential backoff (default=1hr)

Telemetry Metrics

• datadog.logs_agent.http_connectivity_check - Tracks HTTP connectivity check results and TCP fallback events
• datadog.logs_agent.http_connectivity_retry_attempt - Tracks HTTP connectivity retry attempts
• datadog.logs_agent.restart_attempt - Tracks logs agent restart attempts by status and transport type

Motivation

https://datadoghq.atlassian.net/browse/AGNTLOG-392

Describe how you validated your changes

unit tests, local testing (see comment below)

Additional Notes

[agent-platform-auto-pr]

Static quality checks

✅ Please find below the results from static quality gates
Comparison made with ancestor d09c8b3
📊 Static Quality Gates Dashboard

Successful checks

Info

	Quality gate	Delta	On disk size (MiB)	Delta	On wire size (MiB)
✅	agent_deb_amd64	$${+0.02}$$	$${708.11}$$ < $${708.41}$$	$${+0}$$	$${173.9}$$ < $${174.49}$$
✅	agent_deb_amd64_fips	$${+0.02}$$	$${703.41}$$ < $${704.0}$$	$${-0.02}$$	$${172.67}$$ < $${173.75}$$
✅	agent_heroku_amd64	$${+0.02}$$	$${328.72}$$ < $${329.53}$$	$${+0.01}$$	$${87.43}$$ < $${88.45}$$
✅	agent_msi	$${+0.02}$$	$${570.71}$$ < $${982.08}$$	$${+0.04}$$	$${142.71}$$ < $${143.02}$$
✅	agent_rpm_amd64	$${+0.02}$$	$${708.1}$$ < $${708.38}$$	$${-0.03}$$	$${176.44}$$ < $${177.66}$$
✅	agent_rpm_amd64_fips	$${+0.02}$$	$${703.4}$$ < $${703.99}$$	$${+0.01}$$	$${175.94}$$ < $${176.6}$$
✅	agent_rpm_arm64	$${+0.02}$$	$${689.63}$$ < $${693.52}$$	$${-0.02}$$	$${159.93}$$ < $${161.26}$$
✅	agent_rpm_arm64_fips	$${+0.02}$$	$${685.78}$$ < $${688.48}$$	$${-0.01}$$	$${159.42}$$ < $${160.55}$$
✅	agent_suse_amd64	$${+0.02}$$	$${708.1}$$ < $${708.38}$$	$${-0.03}$$	$${176.44}$$ < $${177.66}$$
✅	agent_suse_amd64_fips	$${+0.02}$$	$${703.4}$$ < $${703.99}$$	$${+0.01}$$	$${175.94}$$ < $${176.6}$$
✅	agent_suse_arm64	$${+0.02}$$	$${689.63}$$ < $${693.52}$$	$${-0.02}$$	$${159.93}$$ < $${161.26}$$
✅	agent_suse_arm64_fips	$${+0.02}$$	$${685.78}$$ < $${688.48}$$	$${-0.01}$$	$${159.42}$$ < $${160.55}$$
✅	docker_agent_amd64	$${+0.02}$$	$${769.95}$$ < $${770.72}$$	$${-0}$$	$${261.75}$$ < $${262.45}$$
✅	docker_agent_arm64	$${+0.02}$$	$${776.06}$$ < $${780.2}$$	$${+0.03}$$	$${250.83}$$ < $${252.63}$$
✅	docker_agent_jmx_amd64	$${+0.02}$$	$${960.82}$$ < $${961.6}$$	$${-0.01}$$	$${330.38}$$ < $${331.08}$$
✅	docker_agent_jmx_arm64	$${+0.02}$$	$${955.66}$$ < $${959.8}$$	$${+0.03}$$	$${315.46}$$ < $${317.27}$$
✅	docker_cluster_agent_amd64	$${+0}$$	$${180.7}$$ < $${181.08}$$	$${-0}$$	$${63.85}$$ < $${64.49}$$
✅	docker_cluster_agent_arm64	$${+0}$$	$${196.55}$$ < $${198.49}$$	$${+0}$$	$${60.12}$$ < $${61.17}$$
✅	docker_cws_instrumentation_amd64	$${+0}$$	$${7.13}$$ < $${7.18}$$	$${-0}$$	$${2.99}$$ < $${3.33}$$
✅	docker_cws_instrumentation_arm64	$${+0}$$	$${6.69}$$ < $${6.92}$$	$${+0}$$	$${2.73}$$ < $${3.09}$$
✅	docker_dogstatsd_amd64	$${+0}$$	$${38.78}$$ < $${39.38}$$	$${-0}$$	$${15.01}$$ < $${15.82}$$
✅	docker_dogstatsd_arm64	$${+0}$$	$${37.07}$$ < $${37.94}$$	$${+0}$$	$${14.34}$$ < $${14.83}$$
✅	dogstatsd_deb_amd64	$${+0}$$	$${30.0}$$ < $${30.61}$$	$${+0}$$	$${7.94}$$ < $${8.79}$$
✅	dogstatsd_deb_arm64	$${+0}$$	$${28.15}$$ < $${29.11}$$	$${-0}$$	$${6.81}$$ < $${7.71}$$
✅	dogstatsd_rpm_amd64	$${+0}$$	$${30.0}$$ < $${30.61}$$	$${+0}$$	$${7.95}$$ < $${8.8}$$
✅	dogstatsd_suse_amd64	$${+0}$$	$${30.0}$$ < $${30.61}$$	$${+0}$$	$${7.95}$$ < $${8.8}$$
✅	iot_agent_deb_amd64	$${+0.02}$$	$${42.98}$$ < $${43.29}$$	$${+0}$$	$${11.26}$$ < $${12.04}$$
✅	iot_agent_deb_arm64	$${+0.02}$$	$${40.1}$$ < $${40.92}$$	$${+0.01}$$	$${9.63}$$ < $${10.45}$$
✅	iot_agent_deb_armhf	$${+0.02}$$	$${40.68}$$ < $${41.03}$$	$${+0}$$	$${9.82}$$ < $${10.62}$$
✅	iot_agent_rpm_amd64	$${+0.02}$$	$${42.98}$$ < $${43.29}$$	$${+0}$$	$${11.27}$$ < $${12.06}$$
✅	iot_agent_suse_amd64	$${+0.02}$$	$${42.98}$$ < $${43.29}$$	$${+0}$$	$${11.27}$$ < $${12.06}$$

[cit-pr-commenter]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: 2877c89a-cde2-48ab-bef7-2f0c15b63861

Baseline: b85aa53
Comparison: cf47f97
Diff

Optimization Goals: ✅ No significant changes detected

Experiments ignored for regressions

Regressions in experiments with settings containing erratic: true are ignored.

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
❌	docker_containers_cpu	% cpu utilization	+7.02	[+3.96, +10.08]	1	Logs

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
❌	docker_containers_cpu	% cpu utilization	+7.02	[+3.96, +10.08]	1	Logs
➖	quality_gate_logs	% cpu utilization	+2.17	[+0.67, +3.67]	1	Logs bounds checks dashboard
➖	quality_gate_metrics_logs	memory utilization	+0.66	[+0.45, +0.86]	1	Logs bounds checks dashboard
➖	ddot_metrics_sum_cumulativetodelta_exporter	memory utilization	+0.62	[+0.39, +0.85]	1	Logs
➖	quality_gate_idle	memory utilization	+0.60	[+0.56, +0.65]	1	Logs bounds checks dashboard
➖	quality_gate_idle_all_features	memory utilization	+0.47	[+0.42, +0.51]	1	Logs bounds checks dashboard
➖	file_tree	memory utilization	+0.31	[+0.26, +0.37]	1	Logs
➖	otlp_ingest_metrics	memory utilization	+0.25	[+0.10, +0.40]	1	Logs
➖	ddot_metrics_sum_cumulative	memory utilization	+0.22	[+0.06, +0.38]	1	Logs
➖	uds_dogstatsd_20mb_12k_contexts_20_senders	memory utilization	+0.11	[+0.06, +0.17]	1	Logs
➖	otlp_ingest_logs	memory utilization	+0.05	[-0.06, +0.16]	1	Logs
➖	ddot_metrics_sum_delta	memory utilization	+0.02	[-0.20, +0.24]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.01	[-0.11, +0.13]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.00	[-0.08, +0.08]	1	Logs
➖	ddot_logs	memory utilization	-0.02	[-0.09, +0.05]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	-0.02	[-0.07, +0.03]	1	Logs
➖	uds_dogstatsd_to_api_v3	ingress throughput	-0.02	[-0.15, +0.10]	1	Logs
➖	file_to_blackhole_1000ms_latency	egress throughput	-0.03	[-0.44, +0.38]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	-0.04	[-0.41, +0.33]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	-0.05	[-0.45, +0.36]	1	Logs
➖	docker_containers_memory	memory utilization	-0.27	[-0.35, -0.20]	1	Logs
➖	ddot_metrics	memory utilization	-0.45	[-0.69, -0.21]	1	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	-0.49	[-0.56, -0.41]	1	Logs

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	links
✅	docker_containers_cpu	simple_check_run	10/10
✅	docker_containers_memory	memory_usage	10/10
✅	docker_containers_memory	simple_check_run	10/10
✅	file_to_blackhole_0ms_latency	lost_bytes	10/10
✅	file_to_blackhole_0ms_latency	memory_usage	10/10
✅	file_to_blackhole_1000ms_latency	lost_bytes	10/10
✅	file_to_blackhole_1000ms_latency	memory_usage	10/10
✅	file_to_blackhole_100ms_latency	lost_bytes	10/10
✅	file_to_blackhole_100ms_latency	memory_usage	10/10
✅	file_to_blackhole_500ms_latency	lost_bytes	10/10
✅	file_to_blackhole_500ms_latency	memory_usage	10/10
✅	quality_gate_idle	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_idle	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_idle_all_features	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_logs	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_logs	lost_bytes	10/10	bounds checks dashboard
✅	quality_gate_logs	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	cpu_usage	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	lost_bytes	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	memory_usage	10/10	bounds checks dashboard

Explanation

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

Performance changes are noted in the perf column of each table:

• ✅ = significantly better comparison variant performance
• ❌ = significantly worse comparison variant performance
• ➖ = no significant change in performance

A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".

For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:

1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

2. Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.

3. Its configuration does not mark it "erratic".

CI Pass/Fail Decision

✅ Passed. All Quality Gates passed.

• quality_gate_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check lost_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check lost_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.

[ellataira]

Testing on Local Agent (sha 064be4f)

Hard-coded flow:

• Logs Agent to start on TCP
• Then, smart retry loop inits and restarts with HTTP

SO, to test locally:

• start agent with TCP (it will sleep 10s to allow initial start up, before allowing smartHTTPRestart)
• run log_stream.sh to generate logs before, during, and after HTTP restart
• grep -i "HTTP connectivity\|TEST_SMART_RECOVERY\|Successfully upgraded to HTTP transport\|TEST-TIME" '/var/log/datadog/agent.log'
• check logs received on Datadog

Results:

Logs Generated:

We see the logs all make it to intake "at least once"

Logging:

• Observe the agent restarted: [TEST-TIME] Logs agent restarted with http transport at 2025-11-26T17:47:33Z, and that's where we saw duplicate logs

[ellataira]

moving ^ testing setup with flag TEST_SMART_RECOVERY to parallel branch #43547

[ellataira]

OPEN QUESTIONS / SEEKING THOUGHTS ON …

The currently implemented behavior is:

• If we fail to restart on HTTP, rollback to TCP endpoints to not interfere with processing. Return error so we continue to retry HTTP restart under exponential backoff. Since we know HTTP exists, we want to stick to that.
• What about when we first start up the agent and then immediately try to restart? Some logs won’t be read on the initial start up (on TCP) and therefore wont be passed to the restarted logs agent, when we start and restart in too fast of succession, before the Agent detects all log sources.

... is this what we want?

[gh123man]

This is looking great.
Just some small suggestions

[gh123man]

/merge

[dd-devflow-routing-codex]

View all feedbacks in Devflow UI.

2026-01-06 17:50:49 UTC ℹ️ Start processing command /merge

---

2026-01-06 17:50:57 UTC ℹ️ MergeQueue: waiting for PR to be ready

This pull request is not mergeable according to GitHub. Common reasons include pending required checks, missing approvals, or merge conflicts — but it could also be blocked by other repository rules or settings.
It will be added to the queue as soon as checks pass and/or get approvals.
Note: if you pushed new commits since the last approval, you may need additional approval.
You can remove it from the waiting list with /remove command.

---

2026-01-06 18:40:48 UTC ℹ️ MergeQueue: merge request added to the queue

The expected merge time in main is approximately 1h (p90).

---

2026-01-06 19:49:15 UTC ℹ️ MergeQueue: This merge request was merged