[Refactor] Drop redundant AgentTaskQueue cleanup from the lake alter-job handoff resets

The demotion drain stage (abandonInFlightAgentTasks) empties the whole AgentTaskQueue
before AlterHandler.onStopped() runs the per-job resets, and the activation-side sweep
runs before any daemon can enqueue - so the removeBatchTask / per-task removeTask calls
in the lake jobs' resetTransientStateForHandoff() were no-ops on every reachable path.
Remove them and state the invariant in the comments; the essential part (fresh/nulled
batch so the re-run re-sends or re-dispatches) is unchanged.

Signed-off-by: gengjun-git <gengjun@starrocks.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Author: gengjun-git

fe/fe-core/src/main/java/com/starrocks/alter/LakeRollupJob.java

@@ -133,9 +133,9 @@ protected void resetTransientStateForHandoff() {
         if (jobState == JobState.RUNNING) {
             jobState = JobState.WAITING_TXN;
         }
-        // Same normalization replay() performs: drop queue entries and start from an empty
-        // batch - runWaitingTxnJob appends directly to the field (double-add hazard).
-        AgentTaskQueue.removeBatchTask(rollupBatchTask, TTaskType.ALTER);
+        // Same normalization replay() performs: start from an empty batch - runWaitingTxnJob
+        // appends directly to the field (double-add hazard). No AgentTaskQueue cleanup needed:
+        // the demotion drain (abandonInFlightAgentTasks) already emptied the queue.
         rollupBatchTask = new AgentBatchTask();
         // whereClause deliberately KEPT: this class has no gsonPostProcess restore, so a real
         // reload silently loses the sync-MV filter (pre-existing reload bug); keep the

fe/fe-core/src/main/java/com/starrocks/alter/LakeTableIndexFastPathJobBase.java

@@ -131,14 +131,10 @@ protected void resetTransientStateForHandoff() {
             jobState = JobState.PENDING;
         }
         // runRunningJob dispatches ONLY behind a null guard - a stale batch would silently
-        // suppress the re-send and wedge the job until timeout. Drop queue entries (mirror
-        // cancelImpl) and null the field so the re-run re-dispatches.
-        if (batchTask != null) {
-            for (AgentTask task : batchTask.getAllTasks()) {
-                AgentTaskQueue.removeTask(task.getBackendId(), TTaskType.ALTER, task.getSignature());
-            }
-            batchTask = null;
-        }
+        // suppress the re-send and wedge the job until timeout. Null the field so the re-run
+        // re-dispatches. No AgentTaskQueue cleanup needed: the demotion drain
+        // (abandonInFlightAgentTasks) already emptied the queue before this hook runs.
+        batchTask = null;
     }
 
     protected LakeTableIndexFastPathJobBase(long jobId, JobType jobType, long dbId, long tableId, String tableName,

fe/fe-core/src/main/java/com/starrocks/alter/LakeTableSchemaChangeJob.java

@@ -181,11 +181,11 @@ protected void resetTransientStateForHandoff() {
         if (jobState == JobState.RUNNING) {
             jobState = JobState.WAITING_TXN;
         }
-        // Mirror cancelImpl's queue cleanup, then start from an empty batch: the WAITING_TXN
-        // handler APPENDS to it (double-add hazard), and getInfo dereferences the field, so
-        // fresh-empty rather than null. watershedTxnId/Gtid stay - they are durable with the
+        // Start from an empty batch: the WAITING_TXN handler APPENDS to it (double-add hazard),
+        // and getInfo dereferences the field, so fresh-empty rather than null. No AgentTaskQueue
+        // cleanup needed - the demotion drain (abandonInFlightAgentTasks) already emptied the
+        // queue before this hook runs. watershedTxnId/Gtid stay - they are durable with the
         // WAITING_TXN entry, and runPendingJob reassigns them unconditionally at PENDING.
-        AgentTaskQueue.removeBatchTask(schemaChangeBatchTask, TTaskType.ALTER);
         schemaChangeBatchTask = new AgentBatchTask();
         createReplicaLatch = null;
         waitingCreatingReplica.set(false);