<p>The server aborted a lock wait because another session requested an incompatible exclusive lock.</p>
<p>MySQL Error 1689: ER_LOCK_ABORTED arises when a transaction waiting on a lock is cancelled because another session needs an exclusive lock. Identify and release blocking sessions, shorten transactions, or add proper indexes to resolve and prevent the issue.</p>
Wait on a lock was aborted due to a pending exclusive
MySQL raises error 1689 with SQLSTATE HY000 and message 'Wait on a lock was aborted due to a pending exclusive' when a session waiting on a shared or row-level lock is cancelled because another session needs an exclusive lock on the same resource.
InnoDB solves the conflict by prioritising the exclusive request, rolling back the waiting statement and returning ER_LOCK_ABORTED to the client.
The error surfaces in highly concurrent workloads that mix long running reads with writes, gap locks, or schema changes that require metadata locks.
Long transactions keep row or table locks open longer than necessary, increasing the chance that another connection will request an exclusive lock and force an abort.
Missing or poor indexes push the optimizer to scan and lock more rows than required, again widening the conflict window.
Foreign key cascades and ALTER TABLE operations request metadata locks that immediately abort waiting statements.
First, find the blocking session with SHOW PROCESSLIST or performance_schema and decide whether to wait, tune, or terminate it.
If the blocker is legitimate, modify the waiting query so it acquires fewer or shorter locks: add proper indexes, filter rows, or move to a smaller transaction.
You can also protect critical code paths by using SELECT ... FOR UPDATE NOWAIT or LOCK IN SHARE MODE to fail fast rather than wait indefinitely.
Report queries that read large tables under REPEATABLE READ can be moved to a replica or run with READ COMMITTED isolation to avoid holding long shared locks.
Batch updates that touch many rows should process smaller chunks inside explicit transactions and COMMIT between batches, shortening lock duration.
Schema migrations should be scheduled during low traffic windows or executed with pt-online-schema-change so they do not request long metadata locks.
Keep transactions short and explicit: BEGIN, operation, COMMIT.
Always have covering indexes on foreign key columns and WHERE predicates so writes and reads lock only what they need.
Monitor information_schema.innodb_trx, data_lock_waits, and performance_schema.metadata_locks to spot growing lock queues before they abort.
Error 1205 lock wait timeout exceeded signals that a session waited too long on a lock rather than being aborted immediately.
Error 1213 deadlock found means InnoDB detected a cycle of transactions waiting on each other and rolled one back to break the deadlock.
Error 3572 worker blocked by parallel replication shows similar locking pressure in replicas; analyse the same way.
Galaxy's lightning fast SQL editor surfaces lock waits in real time, letting you inspect SHOW ENGINE INNODB STATUS output, kill sessions, and refactor queries collaboratively. Its AI copilot can rewrite the offending query with better indexes or chunking logic, preventing future ER_LOCK_ABORTED events.
Holding locks for minutes or hours gives other sessions ample time to request exclusive access and force an abort.
Full table scans lock many rows, increasing conflict probability.
Large DML statements lock wide ranges or whole tables.
Cascade deletes/updates acquire extra locks on child tables.
Metadata locks from schema changes abort waiting DML immediately.
Transaction waited longer than innodb_lock_wait_timeout; tune timeout or remove blockers.
Transactions formed a cycle of dependencies; InnoDB rolled one back.
Parallel replication thread blocked on lock; examine data_lock_waits.
MySQL rolls back the current statement; if you use autocommit=0 the whole transaction is usually rolled back, so plan for that possibility.
InnoDB currently prioritises exclusive locks; you cannot change that but you can reduce read duration or use LOCK IN SHARE MODE.
No. A deadlock requires two or more transactions waiting in a cycle. ER_LOCK_ABORTED happens when MySQL preemptively cancels a wait.
No. The server aborts immediately; timeout only controls error 1205. Focus on reducing conflicting locks.
.avif)
.avif){kind=logo}
