mirror of
https://github.com/bytedance/deer-flow.git
synced 2026-11-16 08:20:34 +00:00
* fix(subagents): revalidate durable batch owners before execution * fix(subagents): log batch owner lookup failures safely --------- Co-authored-by: Totoro-qaq <279883115+Totoro-qaq@users.noreply.github.com>
79 lines
4.5 KiB
Markdown
79 lines
4.5 KiB
Markdown
# Native batch dependencies and follow-up
|
|
|
|
Optional fields extend the existing SQL-backed batch service. Omitting them keeps
|
|
independent batch execution. Items remain read-only or idempotent under the existing
|
|
retry/recovery policy.
|
|
|
|
## Dependencies
|
|
|
|
Use `depends_on` to name prerequisite item keys in the same batch. Unknown keys,
|
|
self-references and cycles are rejected before saving. Dependent items consume no
|
|
worker slot until every prerequisite succeeds without a guardrail stop reason.
|
|
Retryable attempts keep dependents waiting; exhausted failures, cancellation or
|
|
terminal capped results fail unstarted descendants. Independent items continue.
|
|
Retrying a failed prerequisite by hand does not requeue the descendants it already
|
|
failed; retry each blocked item as well.
|
|
|
|
Item prompts must name their input/output paths. Ordering does not certify factual
|
|
accuracy or make advisory acceptance mandatory. Use existing result readers; full
|
|
reports are not copied into every dependent's context.
|
|
|
|
## Requested follow-up
|
|
|
|
`continuation` describes remaining work the user already requested. After terminal
|
|
batch results are saved, the lead performs that work in the original conversation
|
|
through ordinary Run admission. Preserve the originating model selection, reasoning,
|
|
limits and admitted retrieval scope. Busy conversations wait. While a question is
|
|
unanswered the claim is kept until its lease expires, so the conversation is not
|
|
reread on every poll.
|
|
|
|
A stable idempotency key reconciles lost admission replies without another Run.
|
|
Pending claims rotate so busy conversations do not starve others. Manual item retries
|
|
hold an unadmitted follow-up but do not re-arm one already dispatched. Without a Goal,
|
|
a follow-up cannot request another automatic follow-up.
|
|
|
|
An active Goal wakes only from its own `external_wait`, consuming one existing
|
|
continuation allowance, once even after recovery. Other blockers, no-progress limits
|
|
and hard stops remain effective. Goal replacement, clearing and thread deletion
|
|
suppress pending work; a failed, cancelled or hard-stopped source Run cannot restart.
|
|
|
|
## Authority, cancellation and recovery
|
|
|
|
Each Gateway batch item rechecks its owner before tool discovery and again after
|
|
acquiring native execution capacity. This applies to the first attempt and to
|
|
recovery after a restart. Missing, disabled or reset accounts cannot start work.
|
|
A changed role rejects the old item; submit a new batch with current permissions.
|
|
Custom roles have no general ordering, so this rule prevents both stale privileges
|
|
after demotion and additional privileges after promotion. OAuth identity comes from
|
|
the current account; serialized authorization attributes are not restored.
|
|
|
|
Explicit local auth-disabled mode and unmapped trusted internal owners retain their
|
|
existing identity semantics. A deleted local account cannot become an unmapped
|
|
internal owner. These checks govern admission, not cancellation of already admitted
|
|
work. Standalone harness hosts with mutable accounts can supply the same
|
|
`resolve_principal` callback without importing the Gateway.
|
|
|
|
Follow-up needs a session-authorized Gateway AgentRuns capability or local auth-disabled
|
|
mode. Internal/PAT callers and unsupported embedded runtimes retain ordinary batches.
|
|
Persist only the original permission ceiling, account token version and auth source;
|
|
no cookie, API key or Python capability. Restore checks current account/permissions,
|
|
conversation ownership and incarnation. Private continuation data is not returned.
|
|
|
|
Batch cancellation revokes the existing SQL execution fence and stops an admitted
|
|
follow-up, including after a lost launch reply. HTTP and tool cancellation of a batch
|
|
with a follow-up also require `runs:cancel`. Trusted internal cleanup retains its existing
|
|
path. Claim ownership/expiry lives only in the existing execution-lease table.
|
|
Completed item results remain available. The Web panel discovers follow-up Runs
|
|
through existing observers; refresh never launches work. Chat Stop remains available.
|
|
|
|
## Upgrade and verification
|
|
|
|
Migration `0052_batch_continuation` adds dependency edges and nullable follow-up data,
|
|
status and Run ID. Existing rows need no backfill. Upgrade all batch workers before
|
|
using dependencies. Do not downgrade with opted-in work outstanding; downgrade drops
|
|
new metadata while retaining items and reports.
|
|
|
|
Focused regressions cover dependency/retry ordering, SQL claims and recovery,
|
|
permission revocation, admission races, Goal limits and page reconnection. PostgreSQL
|
|
coverage uses `TEST_POSTGRES_URI`; browser backend timing is simulated.
|