Language: Python · Sphere: programming · Category: Distributed
In-memory service-discovery registry with heartbeat-based health and TTL.
Use when instances register themselves and clients look them up by name and an optional set of tags: each instance must heartbeat within its TTL to stay discoverable, and a background thread reaps the silent ones. Lookups return only healthy instances. Guarantees (proven by self-test): name+tag lookup is exact (tags matched as a subset); an instance that stops heartbeating disappears once its TTL lapses while a heartbeated peer survives; deregister returns honestly and removes immediately.
When it runs, service registry guarantees len(found) == 2 and found[0].port == 8080; {s.service_id for s in registry.find('web-server')} == {web_id, web2_id}; len(registry.find('database')) == 1 (proven by run).
Checkable constraints:
{s.service_id for s in registry.find('web-server')} == {web_id, web2_id}len(registry.find('database')) == 1registry.find('ghost-service') == [][s.service_id for s in prod_web] == [web_id]registry.find('web-server', {'production', 'frontend'}) != []registry.find('web-server', {'production', 'backend'}) == []len(found) == 2 and found[0].port == 8080registry.heartbeat(web_id) is True
- Green-run: ✓ passes (re-run under the extractor's gate)
- Constraint strength: recovery (truth-pinned)
- Independent oracle: — none yet (green-run candidate; not an axiom under the frozen ruler)
- Peer review: unreviewed
△ AURA Pattern Library — © Reality Optimizer