mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-10-23 09:07:46 +02:00
Cancel blocked queued heartbeat retries
This commit is contained in:
1 parent
616d6700dc
commit
3e8320edda
2 files changed
+254
-2
No files matched your search
@@ -346,6 +346,198 @@ describeEmbeddedPostgres("heartbeat dependency-aware queued run selection", () =
|
||||
expect(blockedWakeRequestCount).toBeGreaterThanOrEqual(2);
|
||||
});
|
||||
|
||||
it("cancels stale queued runs when issue blockers are still unresolved", async () => {
|
||||
const companyId = randomUUID();
|
||||
const agentId = randomUUID();
|
||||
const blockerId = randomUUID();
|
||||
const blockedIssueId = randomUUID();
|
||||
const readyIssueId = randomUUID();
|
||||
const blockedWakeupRequestId = randomUUID();
|
||||
const readyWakeupRequestId = randomUUID();
|
||||
const blockedRunId = randomUUID();
|
||||
const readyRunId = randomUUID();
|
||||
|
||||
await db.insert(companies).values({
|
||||
id: companyId,
|
||||
name: "Paperclip",
|
||||
issuePrefix: `T${companyId.replace(/-/g, "").slice(0, 6).toUpperCase()}`,
|
||||
requireBoardApprovalForNewAgents: false,
|
||||
});
|
||||
await db.insert(agents).values({
|
||||
id: agentId,
|
||||
companyId,
|
||||
name: "QAChecker",
|
||||
role: "qa",
|
||||
status: "active",
|
||||
adapterType: "codex_local",
|
||||
adapterConfig: {},
|
||||
runtimeConfig: {
|
||||
heartbeat: {
|
||||
wakeOnDemand: true,
|
||||
maxConcurrentRuns: 2,
|
||||
},
|
||||
},
|
||||
permissions: {},
|
||||
});
|
||||
await db.insert(issues).values([
|
||||
{
|
||||
id: blockerId,
|
||||
companyId,
|
||||
title: "Security review",
|
||||
status: "blocked",
|
||||
priority: "high",
|
||||
},
|
||||
{
|
||||
id: blockedIssueId,
|
||||
companyId,
|
||||
title: "QA validation",
|
||||
status: "blocked",
|
||||
priority: "medium",
|
||||
assigneeAgentId: agentId,
|
||||
},
|
||||
{
|
||||
id: readyIssueId,
|
||||
companyId,
|
||||
title: "Ready QA task",
|
||||
status: "todo",
|
||||
priority: "low",
|
||||
assigneeAgentId: agentId,
|
||||
},
|
||||
]);
|
||||
await db.insert(issueRelations).values({
|
||||
companyId,
|
||||
issueId: blockerId,
|
||||
relatedIssueId: blockedIssueId,
|
||||
type: "blocks",
|
||||
});
|
||||
await db.insert(agentWakeupRequests).values([
|
||||
{
|
||||
id: blockedWakeupRequestId,
|
||||
companyId,
|
||||
agentId,
|
||||
source: "automation",
|
||||
triggerDetail: "system",
|
||||
reason: "transient_failure_retry",
|
||||
payload: { issueId: blockedIssueId },
|
||||
status: "queued",
|
||||
},
|
||||
{
|
||||
id: readyWakeupRequestId,
|
||||
companyId,
|
||||
agentId,
|
||||
source: "assignment",
|
||||
triggerDetail: "system",
|
||||
reason: "issue_assigned",
|
||||
payload: { issueId: readyIssueId },
|
||||
status: "queued",
|
||||
},
|
||||
]);
|
||||
await db.insert(heartbeatRuns).values([
|
||||
{
|
||||
id: blockedRunId,
|
||||
companyId,
|
||||
agentId,
|
||||
invocationSource: "automation",
|
||||
triggerDetail: "system",
|
||||
status: "queued",
|
||||
wakeupRequestId: blockedWakeupRequestId,
|
||||
contextSnapshot: {
|
||||
issueId: blockedIssueId,
|
||||
wakeReason: "transient_failure_retry",
|
||||
},
|
||||
},
|
||||
{
|
||||
id: readyRunId,
|
||||
companyId,
|
||||
agentId,
|
||||
invocationSource: "assignment",
|
||||
triggerDetail: "system",
|
||||
status: "queued",
|
||||
wakeupRequestId: readyWakeupRequestId,
|
||||
contextSnapshot: {
|
||||
issueId: readyIssueId,
|
||||
wakeReason: "issue_assigned",
|
||||
},
|
||||
},
|
||||
]);
|
||||
await db
|
||||
.update(agentWakeupRequests)
|
||||
.set({ runId: blockedRunId })
|
||||
.where(eq(agentWakeupRequests.id, blockedWakeupRequestId));
|
||||
await db
|
||||
.update(agentWakeupRequests)
|
||||
.set({ runId: readyRunId })
|
||||
.where(eq(agentWakeupRequests.id, readyWakeupRequestId));
|
||||
await db
|
||||
.update(issues)
|
||||
.set({
|
||||
executionRunId: blockedRunId,
|
||||
executionAgentNameKey: "qa-checker",
|
||||
executionLockedAt: new Date(),
|
||||
})
|
||||
.where(eq(issues.id, blockedIssueId));
|
||||
|
||||
await heartbeat.resumeQueuedRuns();
|
||||
|
||||
await waitForCondition(async () => {
|
||||
const run = await db
|
||||
.select({ status: heartbeatRuns.status })
|
||||
.from(heartbeatRuns)
|
||||
.where(eq(heartbeatRuns.id, readyRunId))
|
||||
.then((rows) => rows[0] ?? null);
|
||||
return run?.status === "succeeded";
|
||||
});
|
||||
|
||||
const [blockedRun, blockedWakeup, blockedIssue, readyRun] = await Promise.all([
|
||||
db
|
||||
.select({
|
||||
status: heartbeatRuns.status,
|
||||
errorCode: heartbeatRuns.errorCode,
|
||||
finishedAt: heartbeatRuns.finishedAt,
|
||||
resultJson: heartbeatRuns.resultJson,
|
||||
})
|
||||
.from(heartbeatRuns)
|
||||
.where(eq(heartbeatRuns.id, blockedRunId))
|
||||
.then((rows) => rows[0] ?? null),
|
||||
db
|
||||
.select({
|
||||
status: agentWakeupRequests.status,
|
||||
error: agentWakeupRequests.error,
|
||||
})
|
||||
.from(agentWakeupRequests)
|
||||
.where(eq(agentWakeupRequests.id, blockedWakeupRequestId))
|
||||
.then((rows) => rows[0] ?? null),
|
||||
db
|
||||
.select({
|
||||
executionRunId: issues.executionRunId,
|
||||
executionAgentNameKey: issues.executionAgentNameKey,
|
||||
executionLockedAt: issues.executionLockedAt,
|
||||
})
|
||||
.from(issues)
|
||||
.where(eq(issues.id, blockedIssueId))
|
||||
.then((rows) => rows[0] ?? null),
|
||||
db
|
||||
.select({ status: heartbeatRuns.status })
|
||||
.from(heartbeatRuns)
|
||||
.where(eq(heartbeatRuns.id, readyRunId))
|
||||
.then((rows) => rows[0] ?? null),
|
||||
]);
|
||||
|
||||
expect(blockedRun?.status).toBe("cancelled");
|
||||
expect(blockedRun?.errorCode).toBe("issue_dependencies_blocked");
|
||||
expect(blockedRun?.finishedAt).toBeTruthy();
|
||||
expect(blockedRun?.resultJson).toMatchObject({ stopReason: "issue_dependencies_blocked" });
|
||||
expect(blockedWakeup?.status).toBe("skipped");
|
||||
expect(blockedWakeup?.error).toContain("dependencies are still blocked");
|
||||
expect(blockedIssue).toMatchObject({
|
||||
executionRunId: null,
|
||||
executionAgentNameKey: null,
|
||||
executionLockedAt: null,
|
||||
});
|
||||
expect(readyRun?.status).toBe("succeeded");
|
||||
expect(mockAdapterExecute).toHaveBeenCalledTimes(1);
|
||||
});
|
||||
|
||||
it("suppresses normal wakeups while allowing comment interaction wakes under a pause hold", async () => {
|
||||
const companyId = randomUUID();
|
||||
const agentId = randomUUID();
|
||||
|
||||
@@ -3694,9 +3694,11 @@ export function heartbeatService(db: Db) {
|
||||
}
|
||||
|
||||
const dependencyReadiness = await issuesSvc.listDependencyReadiness(run.companyId, [issueId]);
|
||||
const unresolvedBlockerCount = dependencyReadiness.get(issueId)?.unresolvedBlockerCount ?? 0;
|
||||
const readiness = dependencyReadiness.get(issueId);
|
||||
const unresolvedBlockerCount = readiness?.unresolvedBlockerCount ?? 0;
|
||||
if (unresolvedBlockerCount > 0 && !allowsIssueInteractionWake(context)) {
|
||||
logger.debug({ runId: run.id, issueId, unresolvedBlockerCount }, "claimQueuedRun: skipping blocked run");
|
||||
await cancelQueuedRunForBlockedDependencies(run, issueId, readiness?.unresolvedBlockerIssueIds ?? []);
|
||||
logger.info({ runId: run.id, issueId, unresolvedBlockerCount }, "claimQueuedRun: cancelled blocked queued run");
|
||||
return null;
|
||||
}
|
||||
}
|
||||
@@ -3758,6 +3760,64 @@ export function heartbeatService(db: Db) {
|
||||
return claimed;
|
||||
}
|
||||
|
||||
async function cancelQueuedRunForBlockedDependencies(
|
||||
run: typeof heartbeatRuns.$inferSelect,
|
||||
issueId: string,
|
||||
unresolvedBlockerIssueIds: string[],
|
||||
) {
|
||||
const now = new Date();
|
||||
const reason =
|
||||
"Cancelled because issue dependencies are still blocked; Paperclip will wake the assignee when blockers resolve";
|
||||
const cancelled = await setRunStatus(run.id, "cancelled", {
|
||||
finishedAt: now,
|
||||
error: reason,
|
||||
errorCode: "issue_dependencies_blocked",
|
||||
resultJson: {
|
||||
...parseObject(run.resultJson),
|
||||
stopReason: "issue_dependencies_blocked",
|
||||
effectiveTimeoutSec: 0,
|
||||
timeoutConfigured: false,
|
||||
timeoutSource: "dependency_gate",
|
||||
timeoutFired: false,
|
||||
},
|
||||
});
|
||||
if (!cancelled) return null;
|
||||
|
||||
await setWakeupStatus(run.wakeupRequestId, "skipped", {
|
||||
finishedAt: now,
|
||||
error: reason,
|
||||
});
|
||||
|
||||
await db
|
||||
.update(issues)
|
||||
.set({
|
||||
executionRunId: null,
|
||||
executionAgentNameKey: null,
|
||||
executionLockedAt: null,
|
||||
updatedAt: now,
|
||||
})
|
||||
.where(
|
||||
and(
|
||||
eq(issues.companyId, run.companyId),
|
||||
eq(issues.id, issueId),
|
||||
eq(issues.executionRunId, run.id),
|
||||
),
|
||||
);
|
||||
|
||||
await appendRunEvent(cancelled, await nextRunEventSeq(cancelled.id), {
|
||||
eventType: "lifecycle",
|
||||
stream: "system",
|
||||
level: "warn",
|
||||
message: reason,
|
||||
payload: {
|
||||
issueId,
|
||||
unresolvedBlockerIssueIds,
|
||||
},
|
||||
});
|
||||
|
||||
return cancelled;
|
||||
}
|
||||
|
||||
async function finalizeAgentStatus(
|
||||
agentId: string,
|
||||
outcome: "succeeded" | "failed" | "cancelled" | "timed_out",
|
||||
|
||||
Reference in new issue
Block a user