fix(aws service): use http client when building assume role for AccessKey

[StephenWakely]

Fixes #20282

This fixes the error that we received when using both access_key_id and assume_role together. The builder used in that situation should have had a custom http client provided so it could use our TLS and proxy settings.

Note: I have tested this by running locally with that authentication scheme in the config. This gave me the same error in the issue. I fixed the bug and ran again. I now get the error STS refused to grant assume role, which is expected because I have not set up a valid client in AWS. I am pretty confident that this fixes the issue, but have not tested 100% to the end to save time having to setup a fairly complex authentication scheme in AWS.

[datadog-vectordotdev]

Datadog Report

Branch report: stephen/aws_assume_role_provider
Commit report: 255e472
Test service: vector

✅ 0 Failed, 87 Passed, 0 Skipped, 13m 31.66s Wall Time
❄️ 3 New Flaky

New Flaky Tests (3)

• sources::aws_s3::integration_tests::s3_process_message - vector - Last Failure

  Expand for error

   Test has failed
  

• sources::aws_s3::integration_tests::s3_process_message_spaces - vector - Last Failure

  Expand for error

   Test has failed
  

• sources::aws_s3::integration_tests::s3_process_message_with_log_namespace - vector - Last Failure

  Expand for error

   Test has failed
  

[jszwedko]

👍 good catch

[jszwedko]

Do we need any similar changes to the AwsAuthentication::File arm? It seems like it may also not be using the custom HTTP client.

[StephenWakely]

Do we need any similar changes to the AwsAuthentication::File arm? It seems like it may also not be using the custom HTTP client.

Yikes! I assumed it wouldn't because I thought it was just loading from a file, but a quick test shows that that was a bad assumption. Good catch!

[github-actions]

Regression Detector Results

Run ID: 2c04cde2-4635-4977-9292-f573882f1895
Baseline: fae2ebf
Comparison: 1b0bdcf
Total CPUs: 7

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

No significant changes in experiment optimization goals

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

There were no significant changes in experiment optimization goals at this confidence level and effect size tolerance.

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI
➖	syslog_log2metric_splunk_hec_metrics	ingress throughput	+3.25	[+3.07, +3.43]
➖	http_elasticsearch	ingress throughput	+1.73	[+1.64, +1.82]
➖	syslog_regex_logs2metric_ddmetrics	ingress throughput	+1.49	[+1.31, +1.68]
➖	file_to_blackhole	egress throughput	+1.27	[-1.08, +3.61]
➖	datadog_agent_remap_blackhole	ingress throughput	+0.95	[+0.82, +1.08]
➖	syslog_loki	ingress throughput	+0.64	[+0.53, +0.74]
➖	http_to_s3	ingress throughput	+0.63	[+0.35, +0.91]
➖	datadog_agent_remap_datadog_logs	ingress throughput	+0.62	[+0.50, +0.74]
➖	http_to_http_acks	ingress throughput	+0.41	[-0.94, +1.76]
➖	splunk_hec_route_s3	ingress throughput	+0.39	[-0.08, +0.86]
➖	http_to_http_noack	ingress throughput	+0.19	[+0.10, +0.29]
➖	syslog_log2metric_humio_metrics	ingress throughput	+0.14	[-0.01, +0.30]
➖	otlp_grpc_to_blackhole	ingress throughput	+0.12	[+0.03, +0.21]
➖	http_to_http_json	ingress throughput	+0.04	[-0.04, +0.12]
➖	syslog_log2metric_tag_cardinality_limit_blackhole	ingress throughput	+0.03	[-0.13, +0.19]
➖	splunk_hec_to_splunk_hec_logs_acks	ingress throughput	+0.00	[-0.14, +0.14]
➖	splunk_hec_indexer_ack_blackhole	ingress throughput	-0.01	[-0.15, +0.14]
➖	splunk_hec_to_splunk_hec_logs_noack	ingress throughput	-0.05	[-0.16, +0.07]
➖	enterprise_http_to_http	ingress throughput	-0.11	[-0.19, -0.03]
➖	syslog_splunk_hec_logs	ingress throughput	-0.45	[-0.54, -0.37]
➖	datadog_agent_remap_datadog_logs_acks	ingress throughput	-0.64	[-0.73, -0.55]
➖	otlp_http_to_blackhole	ingress throughput	-0.72	[-0.84, -0.60]
➖	datadog_agent_remap_blackhole_acks	ingress throughput	-0.76	[-0.91, -0.62]
➖	http_text_to_http_json	ingress throughput	-2.60	[-2.75, -2.44]
➖	socket_to_socket_blackhole	ingress throughput	-3.15	[-3.24, -3.06]
➖	syslog_humio_logs	ingress throughput	-3.19	[-3.41, -2.96]
➖	fluent_elasticsearch	ingress throughput	-3.55	[-4.02, -3.07]

Explanation

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".