Snapshot Parsing

Reminder: the snapshot log group (/aws/service-events/{service}) must be encrypted at rest with a KMS CMK (aws logs associate-kms-key) before capturing — captured snapshots may contain credentials, PII, or secrets. See Security Considerations in dynamic-instrumentation.md.

Retrieval commands

Both require a host with python3 + boto3. Region resolves as --region flag > AWS_REGION > AWS_DEFAULT_REGION > us-east-1 default (the same precedence as di_instrumentation.py) — it MUST be the same region the breakpoint was created in, or searches return empty even when the breakpoint is ACTIVE. Pass arguments via --json-file (or --json - on stdin) so values stay off the shell command line.

• Discover the snapshot structure first (Step 3 rule — always do this before searching). Write the arguments to a file, then:

  bashCopy

  # args.json:
  # {"service": "<svc>", "environment": "<env>",
  #  "location_hash": "<16-hex>", "status_timestamp": "<ACTIVE-event-ISO8601>"}
  python3 scripts/cloudwatch/di_snapshots.py sample --json-file args.json

  Returns one nearby snapshot as JSON plus per-attribute field_documentation. Read the field paths from this sample — they are authoritative; do not rely on canned paths that may be stale.

• Search a batch near a status-event timestamp, narrowing with custom_filters when you can name the target:

  bashCopy

  # args.json:
  # {"service": "<svc>", "environment": "<env>",
  #  "location_hash": "<16-hex>", "status_timestamp": "<ISO8601>",
  #  "limit": 20, "custom_filters": ["..."]}
  python3 scripts/cloudwatch/di_snapshots.py search --json-file args.json --out /tmp/snaps.json

  custom_filters are raw Logs Insights fragments appended with and; an unbalanced double quote is rejected. --out writes the result with owner-only (0600) permissions because snapshots may contain PII/secrets.

--print-contract lists both ops and their exact argument schema.

Parsing the saved output (never hand-transcribe; never cat a large file into context)

Large results are written to a file (use --out, or redirect stdout). Parse the file with jq/python and extract only the fields you need. Do not retype values you see in tool output into a script literal — a single mistyped orderId/paymentRef silently corrupts the aggregation.

Encryption at rest for the saved file. --out already restricts the file to owner-only (0600), but the snapshot may still contain credentials/PII. Write it only to a private location on an encrypted volume (e.g. an encrypted EBS volume or encrypted tmpfs), avoid world-readable shared temp directories, and delete it as soon as analysis is done.

The retrieval output is JSON. Snapshot records are under results[*], each with an @message that is itself a JSON string — json.loads it again to reach body.captures.*. The parser already extracts the common debugging fields; key ones from a parsed snapshot:

Java Map/HashMap values appear as key/value entries (not flat fields); raise object depth / collection width if map contents are truncated.

Write the jq/python against the actual field paths from your live sample snapshot, group by a domain identifier (e.g. orderId), and surface anomalies (duplicates, outliers). When combining results from multiple queries, deduplicate by snapshot id before aggregating.

After analysis, do not retain the saved snapshot file — it may contain PII/secrets.