feat: add opt-in tracemalloc Python heap metrics (AI-Triage PR)

[devin-ai-integration]

This PR targets the following PR:

• feat: add memory & resource metrics to Python CDK via DogStatsD (Tier 1) #931

---

feat: add opt-in tracemalloc Python heap metrics (AI-Triage PR)

Summary

Re-introduces get_python_heap_bytes() (removed from this branch in commit c8b0197 due to reviewer concerns about unconditional overhead) as an opt-in feature gated behind the CDK_TRACEMALLOC_ENABLED environment variable.

When set to a truthy value (1, true, yes), tracemalloc.start() is called lazily on first metric emission and the cdk.memory.python_heap_bytes gauge is emitted alongside existing container memory metrics. When unset (the default), the function is a no-op returning None — no tracing overhead, no metric emitted.

Files changed:

• airbyte_cdk/metrics/memory.py — adds _is_tracemalloc_enabled() and get_python_heap_bytes()
• airbyte_cdk/metrics/__init__.py — imports new function, adds METRIC_MEMORY_PYTHON_HEAP_BYTES constant, emits the gauge in emit_memory_metrics()
• unit_tests/metrics/test_memory.py — 7 new tests covering enabled/disabled/case-insensitive env var behavior
• unit_tests/metrics/test_metrics_client.py — 2 new tests verifying MetricsClient emits/skips the heap metric

Resolves https://github.com/airbytehq/airbyte-internal-issues/issues/15952:

• https://github.com/airbytehq/airbyte-internal-issues/issues/15952

Review & Testing Checklist for Human

• Env var naming: Confirm CDK_TRACEMALLOC_ENABLED follows the team's naming conventions for CDK env vars (vs. e.g. AIRBYTE_CDK_TRACEMALLOC_ENABLED).
• Top-level import tracemalloc: The stdlib module is imported at module level per Python coding standards, even though the original (removed) implementation used a lazy inline import. Verify this is acceptable — the import itself is cheap but it does load the module unconditionally.
• No test for RuntimeError path: The except RuntimeError branch in get_python_heap_bytes() (when tracemalloc.start() fails) has no dedicated test. Assess if this gap matters.
• Base branch: This PR targets devin/1772669295-add-memory-metrics-module (PR feat: add memory & resource metrics to Python CDK via DogStatsD (Tier 1) #931), not main. Ensure it is rebased/merged correctly after feat: add memory & resource metrics to Python CDK via DogStatsD (Tier 1) #931 lands.

Suggested manual test plan:

1. Check out this branch, set CDK_TRACEMALLOC_ENABLED=true, and run a short connector sync — verify cdk.memory.python_heap_bytes appears in DogStatsD output.
2. Unset the env var and repeat — verify no tracemalloc overhead and no heap metric emitted.

Notes

• Devin session
• Requested by: bot_apk (apk@cognition.ai)
• All 42 metrics tests pass locally (33 existing + 9 new)

Updates since last revision

• Resolved merge conflict with base branch (base added import sys for macOS ru_maxrss handling; this branch adds import tracemalloc; both are now present).

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring

[github-actions]

PyTest Results (Fast)

3 911 tests  +9   3 899 ✅ +9   6m 14s ⏱️ -31s
    1 suites ±0      12 💤 ±0 
    1 files   ±0       0 ❌ ±0 

Results for commit a58279d. ± Comparison against base commit b208bd0.

[github-actions]

PyTest Results (Full)

3 914 tests  +9   3 902 ✅ +9   11m 29s ⏱️ +15s
    1 suites ±0      12 💤 ±0 
    1 files   ±0       0 ❌ ±0 

Results for commit a58279d. ± Comparison against base commit b208bd0.