agentleFS
all-source-osall-source-os/all-source/.claude/skills/chronos-durability/SKILL.md

chronos-durability

Skill4 starsChanged yesterdayApache-2.0View on GitHub

Test AllSource Core event durability by writing events, restarting the container, and verifying all events survive. Proves WAL + Parquet persistence guarantees. Triggers on "test durability", "durability test", "restart test", "verify persistence", "crash recovery test", "prove events survive restart".

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run the durability test script:

    line 15
  2. Distinctive:

    Use appropriate flags based on context:

    line 27

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Chronos Durability Test

Proves that events written to AllSource Core survive container restarts — the core durability guarantee of WAL + Parquet.

When invoked:

  1. Run the durability test script:
    bash tooling/durability-test/test-durability.sh
  2. Analyze the output and report:
    • Whether all events were written successfully
    • Pre-restart storage state (WAL, Parquet if available)
    • Recovery time after restart
    • Whether all events survived the restart (the money test)
    • Post-restart storage state comparison
  3. Use appropriate flags based on context:
    • --target docker (default) — local Docker stack (Core on port 3280)
    • --target compose — Docker compose with internal ports (Core on port 3900)
    • --target fly — Fly.io deployment
    • --count N — number of events to write (default: 10)
    • --skip-restart — write and verify only, no restart cycle
    • --json — machine-readable output
  4. For quick smoke tests (no restart needed):
    bash tooling/durability-test/test-durability.sh --skip-restart

Test Phases

PhaseWhat's Checked
Pre-flightCore reachable, docker/fly CLI available, container exists
BaselineCurrent event count via /api/v1/stats
WriteIngest N events with unique durability-test-{timestamp} entity
Verify WriteQuery back all events, confirm count matches
Storage CheckWAL status, storage stats, deep health (informational, may skip)
Restartdocker restart or fly machines restart
RecoveryPoll /health every 2s until back (timeout 60s)
Verify DurabilityQuery same events — ALL N must be present
InvestigationPost-restart stats, WAL status, deep health comparison

Common Failure Patterns

  • Events lost after restart: Volume not mounted or WAL disabled. Check ALLSOURCE_WAL_ENABLED=true and Docker volume mounts.
  • Partial recovery: WAL replay incomplete. May indicate crash during fsync interval. Check ALLSOURCE_WAL_SYNC_INTERVAL_MS.
  • Recovery timeout: Container takes too long to start. Check Docker logs for WAL replay errors.
  • Ops endpoints skip: /api/v1/ops/* endpoints may not exist in all Core versions. This is informational only, not a failure.

Architecture Note

This test writes directly to Core (POST /api/v1/events) to eliminate the Query Service as a variable. Core IS the database — it has WAL (CRC32, fsync), Parquet (Snappy), and DashMap. Events are durable by design.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.