Lock Management
In DataArts Fabric SQL, LakeFormation central locks are used to ensure correctness when multiple compute nodes concurrently write to the same table. This section introduces parameters primarily governing the behavior of LakeFormation central locks.
fabricsql_lflock_wait_timeout
Definition: Controls the maximum wait time for acquiring a lock. An error will be reported if the wait time for requesting a lock exceeds the set value.
Range: an integer ranging from 0 to INT_MAX. The unit is milliseconds (ms).
- If the value of this parameter equals 0, it means at most one attempt is allowed. If the attempt fails, an error is reported directly without retry.
Default Value: 30s
fabricsql_lflock_heartbeat_timeout
Definition: Sets the maximum timeout duration for lock heartbeat failure.
- To maintain the liveliness of LakeFormation central locks, the system continuously sends lock heartbeats to LakeFormation while holding the lock. Upon failure to send a lock heartbeat, the system does not immediately report an error but keeps retrying until reaching the maximum timeout, after which an error is reported if unsuccessful.
- LakeFormation's automatic lock release upon missing a heartbeat is 60s, with the system normally sending heartbeats every 30s. Thus, setting this parameter below 30s causes immediate error reporting on the first heartbeat failure, whereas above 60s risks losing the lock during prolonged disconnection from LakeFormation. If you want to avoid lost locks under extended disconnections, you should either set this parameter below 60s or ensure no concurrent writes occur within the configured interval.
Range: an integer ranging from 0 to INT_MAX. The unit is milliseconds (ms).
Default Value: 5 min
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot