From software problem to a clear next step.
Choose a topic and what you need to do next. Add a search phrase to find a specific problem.
Browse 20 topics
Explore the playbooks
1500 guides · Page 35 of 60
-
Triage long-job time limits by phase and cleanup state
Locate a long-job time-limit failure by comparing deadline configuration, current phase, cleanup behavior, commit boundary, worker exit, and terminal state.
-
Prioritize long-job timeouts by recovery and side-effect risk
Rank time-limit work by partial durable effects, unsafe retries, abandoned resources, blocked dependents, and confidence in the current cleanup and terminal-state contract.
-
Investigate long-job timeouts around the commit boundary
Build a reproducible timeout brief by testing deadlines before, during, and after a known commit or cleanup boundary and tracing the resulting state.
-
Verify long-job limits below, at, and beyond the deadline
Check long jobs that finish before the limit, hit it during work, and cross it during cleanup, confirming durable effects, terminal states, and retry safety.
-
Document phase-specific time limits and cleanup behavior
Preserve lessons about deadline calculation, phase budgets, commit ownership, cleanup idempotency, retry states, and fixtures for future long-job changes.
-
Triage scheduled timezone drift from local rule to UTC run
Locate a scheduled job timezone shift by comparing configured zone, intended local time, computed due time, host clock, actual start, and calendar boundary.
-
Prioritize scheduled timezone drift by missed calendar consequence
Rank timezone scheduling work by missed commitments, duplicate or early runs, local calendar boundaries affected, and confidence in the stored rule and due timestamp.
-
Investigate scheduled timezone drift across a calendar boundary
Build a reproducible scheduling brief by replaying one recurrence before and after an offset transition and tracing local rule, UTC due time, enqueue, and start.
-
Verify scheduled runs before and after a timezone transition
Check ordinary dates and both sides of the relevant calendar boundary, confirming local display, UTC due time, actual start, and occurrence identity remain coherent.
-
Keep an explicit timezone contract for recurring jobs
Preserve schedule timezone, local rule, UTC calculation, occurrence audit, and ordinary and boundary fixtures for future calendar and scheduler changes.
-
Triage a missed scheduled run across every handoff
Locate a missed scheduling window by tracing due, enqueue, ready, claim, start, expiration, worker availability, and terminal outcome for one occurrence.
-
Prioritize missed scheduled runs by commitment and recovery risk
Rank missed-window work by the action promised, lateness consequence, duplicate replay risk, neighboring-run evidence, and whether a safe recovery choice is still available.
-
Investigate a missed scheduled occurrence from due time onward
Build a reproducible missed-window brief by comparing one missing occurrence with neighboring successes and exercising the scheduler-to-queue boundary.
-
Verify scheduled runs in available and unavailable worker windows
Check a due run with workers available, a delayed worker, and an explicit recovery decision, confirming occurrence identity, handoffs, terminal state, and no duplicate replay.
-
Document missed-window signals and scheduled-run recovery
Preserve due-to-start handoffs, occurrence identity, worker availability evidence, replay rules, and fixtures for detecting and recovering missed scheduled work.
-
Triage overlapping scheduled runs with IDs and lock state
Locate unexpected schedule overlap by comparing run IDs, due intervals, duration, lock or lease state, worker claims, and the configured overlap policy.
-
Prioritize scheduled overlap by duplicate effect risk
Rank overlapping-run work by duplicate writes, competing output ownership, resource contention, downstream confusion, and confidence that the overlap violates the stated policy.
-
Investigate scheduled overlap with a deliberately slow run
Build a reproducible overlap brief by spanning one schedule interval and tracing due events, lock or lease decisions, worker claims, and output ownership.
-
Verify scheduled runs that finish before and after the interval
Check schedule runs that finish before the next due time and those that span it, confirming the chosen skip, queue, lock, or concurrency policy and durable outputs.
-
Document the scheduled-run overlap policy and lease boundary
Preserve overlap rules, run identity, lock or lease ownership, output reconciliation, and short and spanning fixtures for future schedule changes.
-
Triage expired job output from creation to reader request
Locate why completed job output is unavailable by comparing job terminal state, output key, creation time, expiry policy, read request, and any rebuild path.
-
Prioritize job output expiry by lost-work consequence
Rank output retention work by completed work that cannot be consumed, downstream decisions blocked, rebuild cost, reader delay, and confidence in the expiry and recovery contract.
-
Investigate job output expiry with delayed reader fixtures
Build a reproducible expiry brief by following output creation, retention, cache or storage read, and rebuild behavior inside and beyond the allowed window.
-
Verify job output reads inside and after the retention window
Check immediate, boundary, and expired output reads, confirming status messages, metadata, rebuild or empty states, and that readers cannot mistake missing work for success.
-
Document job output retention and reader recovery
Preserve output identity, retention timing, reader states, rebuild semantics, and fixtures for completed job results that may expire before delayed use.