Skip to content

tutorial.md

Latest commit

 

History

History
88 lines (60 loc) · 2.36 KB

tutorial.md

File metadata and controls

88 lines (60 loc) · 2.36 KB
Raw

Tutorial

These examples show how to use StorageOps without exposing credentials or granting cloud account access.

1. 429 SlowDown

storageops --print 's5cmd sync s3://bucket/data ./data fails with 429 SlowDown'

Good evidence to include:

  • tool and version,
  • command shape,
  • concurrency settings,
  • object count and average size,
  • whether errors cluster by prefix.

StorageOps should route to storageops-performance-diagnosis.

2. SignatureDoesNotMatch

storageops --print @error-response.xml 'why does this S3-compatible endpoint reject the signature?'

For saved XML/debug traces, use the helper:

python3 skills/storageops-s3-protocol-compatibility/scripts/parse_sigv4_error.py \
  error-response.xml --json

StorageOps should inspect credential scope, region, signed headers, payload hash, and canonical request shape.

3. Endpoint Timeout

storageops --print 'EC2 in private subnet times out connecting to s3.us-east-1.amazonaws.com'

If the user explicitly wants an active check from this machine:

python3 skills/storageops-network-endpoint-access/scripts/endpoint_reachability_test.py \
  https://s3.us-east-1.amazonaws.com --skip-http

StorageOps should distinguish DNS, TCP, TLS, and application-layer failures.

4. BadDigest On Upload

storageops --print @server-error.xml @upload-debug.txt 'why does this PUT return BadDigest?'

Do not re-send the write just to trace it. Provide the server error body and the client's debug dump; if the issue may be a gzip/payload-hash mismatch, use:

python3 skills/storageops-s3-protocol-compatibility/scripts/check_payload_hash.py \
  --raw-file out.json --declared-sha256 <hex> --content-encoding gzip --json

StorageOps should diagnose from the failed request evidence and keep live HTTP trace use read-only.

5. Access Log Investigation

storageops --print @access-log-sample.txt 'identify who caused this 503 spike'

StorageOps should route to access-log analysis, aggregate by requester and operation, then escalate to performance only after attribution.

6. Generate a Report

storageops --print @diagnosis-notes.md 'turn this into a customer-facing report'

StorageOps should route to storageops-evidence-reporting and preserve:

  • summary,
  • key evidence,
  • root cause,
  • recommendations,
  • limitations and confidence.