* feat(sandbox): pluggable per-engine registry (postgres/redis/mongo) Replaces the hardcoded postgres+redis branches in the provisioner and the env emitter with a registry of SandboxEngine specs (image, run args, readiness probe, connection, ROBOCO_TEST_* env) in a pure low module (roboco/models/sandbox.py). VALID_SANDBOX_SERVICES is derived from the registry — single source of truth — and the provisioner + orchestrator iterate it, so adding an engine is one class + one registry line, not another branch. Adds a mongo:8-alpine engine (ROBOCO_TEST_MONGO_*) as the third service alongside postgres/redis. Also fixes the cold-pull loop that stranded v0.19.0 board agents with empty error strings: docker run pulled inline under a 20s deadline, so a NAS cold pull was killed, cancelled, and re-pulled from scratch forever. _ensure_image now inspects + pulls (300s) before run; provisioning errors log type+message so a bare TimeoutError no longer shows as "". Panel edit-project dialog: postgres/redis toggles -> a Set<string> multi-select driven by a SANDBOX_SERVICES catalog, so new engines appear in the UI by adding to the catalog. Tests: engine parity (allowlist==registry, unique slugs/images, no None leak in env, SandboxInfo aggregates every engine), mongo provision + env injection, plus the existing postgres/redis provision/env/spawn/janitor suite updated to the registry shape. 821 unit / 5 skip green; ruff + mypy (360 files) clean. * docs(sandbox): reflect pluggable engine registry + mongo across docs CHANGELOG (0.19.0): Added entry for the pluggable sandbox engine registry (postgres/redis/mongo) + Fixed entry for the cold-pull loop/empty-error strand that boarded v0.19.0 board agents. docs/map (9 files): sandbox subsystem blurbs, SandboxProvisioner rows, _maybe_provision_sandbox/_append_sandbox_env rows, feature-flag rows, the migration-057 row + v0.17.0 delta, and the models.md VALID_SANDBOX_SERVICES note — all retitled to DB/Redis/Mongo via the engine registry (roboco/models/sandbox.py), with the one-class-one-line extension story and the _ensure_image cold-pull fix. Production-network (roboco_data) lines left as postgres+redis — mongo is sandbox-only, not a prod service. docs/rag (3 files): sandbox-db.md rewritten around the registry (engine list, generic _provision_engine, image pre-pull, ROBOCO_TEST_DB_*/REDIS_*/MONGO_* incl. MONGO_AUTH_DB=admin, single emit_env); config-reference sandbox flag row + subsection retitled; db-network-isolation framing broadened to postgres/redis/mongo. preconditions-and-rejections left untouched (its hit was an unrelated gateway see-also link). * test(e2e): harden umbrella close terminal reads with bounded wait-for-state The MegaTask umbrella close test flaked once on CI (ceo-approve returned 200 but the re-fetch saw awaiting_pm_review) then passed on re-run. The production path is deterministic: complete -> main_pm_complete -> submit_pm_review -> escalate_to_ceo -> ceo_approve -> commit, all on one session, all awaited; the fire-and-forget completion hooks are isolated (own session, best-effort, never touch task.status or the request session). 20 local runs could not reproduce it. The one real surface is the read pattern: the e2e stack commits on the uvicorn thread's loop and reads via a separate loop (run_db -> asyncio.run with a fresh engine), so a terminal single point-read can race a still-draining completion hook on a contended runner. Replace the two terminal point-reads with a bounded wait_for_status poll. Strictly better than a one-shot read: absorbs the transient, and a genuine state bug still surfaces via the timeout branch asserting against the last-read state. --------- Co-authored-by: Renn F <rennf93@users.noreply.github.com>
2.5 KiB
DB Network Isolation
What It Is
A compose-topology hardening: two user-defined Docker bridges instead of one. roboco_default carries the agent mesh (panel, nginx, ollama, every spawned agent container, and their sandbox sidecars — postgres / redis / mongo via the engine registry in roboco/models/sandbox.py; see docs/rag/architecture/sandbox-db.md). roboco_data carries only postgres + redis. The orchestrator is the sole multi-homed service (both networks) — every agent container structurally cannot resolve or TCP-reach roboco-postgres:5432 / roboco-redis:6379 at all. This matters because redis has no auth in this deployment: network membership is the containment, not a password.
Enable/Disable
| Variable | Default | Effect |
|---|---|---|
ROBOCO_DB_NETWORK_ISOLATED |
false |
Not a panel feature flag — it must travel with the compose file's networks: stanzas (it describes topology, not a runtime-toggleable behavior), so it is deliberately absent from roboco/services/settings.py's FEATURE_FLAGS. Set true only by the compose files that actually carry the two-bridge topology. |
What flipping it changes
ROBOCO_DB_NETWORK_ISOLATED=true suppresses the legacy _append_gate_env prod-creds injection (roboco/runtime/orchestrator.py) — the one that would otherwise hand an agent ROBOCO_TEST_DB_HOST=roboco-postgres credentials for a host it cannot reach. A connect timeout is worse than no credentials at all (the test suite's DB-reachability check skips cleanly on a fast refusal, but hangs on a dead-end timeout), so the flag makes that injection a no-op rather than let it happen and fail slow. Projects that need a real DB for their gate opt into the sandboxed dev DB/Redis/Mongo instead (docs/rag/architecture/sandbox-db.md) — sandbox replaces, never coexists with, the prod-creds path.
What is unaffected
Agent↔agent A2A (:9000), orchestrator→agent SDK polls, MCP→orchestrator (:8000), and host-published ports (15432/16379/11435 in this topology) are unaffected — those don't route through roboco_data. docker exec / docker inspect paths ride the daemon socket, not the network, so they're untouched too.
Related
docs/rag/architecture/config-reference.md— env var tabledocs/rag/architecture/sandbox-db.md— the alternative for a DB-needing project under this topologydocs/rag/architecture/cloud-auth.md— a separate, unrelated hardening that ships alongside this on the NAS composes