Merge pull request #10298 from knz/fix-retry-release

sql: release discarded results from retried queries

pkg/sql/executor.go

@@ -107,7 +107,12 @@ type StatementResults struct {
 
 // Close ensures that the resources claimed by the results are released.
 func (s *StatementResults) Close() {
-	for _, r := range s.ResultList {
+	s.ResultList.Close()
+}
+
+// Close ensures that the resources claimed by the results are released.
+func (rl ResultList) Close() {
+	for _, r := range rl {
 		r.Close()
 	}
 }
@@ -564,6 +569,10 @@ func (e *Executor) execRequest(session *Session, sql string, copymsg copyMsg) St
 			}
 
 			var err error
+			if results != nil {
+				// Some results were produced by a previous attempt. Discard them.
+				ResultList(results).Close()
+			}
 			results, remainingStmts, err = runTxnAttempt(e, planMaker, origState, txnState, opt, stmtsToExec)
 
 			// TODO(andrei): Until #7881 fixed.

pkg/sql/txn_restart_test.go

@@ -505,9 +505,21 @@ CREATE TABLE t.test (k INT PRIMARY KEY, v TEXT, t DECIMAL);
 	// of txns (statements batched together with the BEGIN stmt) - are retried.
 	// We also exercise the SQL cluster logical timestamp in here, because
 	// this must be properly propagated across retries.
+	//
+	// The SELECT within the transaction also checks that discarded
+	// intermediate result sets are properly released: the result set it
+	// produces is accounted for by the session monitor, and if it is
+	// not properly released upon a retry the monitor will cause the
+	// server to panic (and thus the test to fail) when the connection
+	// is closed.
+	//
+	// TODO(knz) This test can be made more robust by exposing the
+	// current allocation count in monitor and checking that it has the
+	// same value at the beginning of each retry.
 	if _, err := sqlDB.Exec(`
 INSERT INTO t.test(k, v, t) VALUES (1, 'boulanger', cluster_logical_timestamp());
 BEGIN;
+SELECT * FROM t.test;
 INSERT INTO t.test(k, v, t) VALUES (2, 'dromedary', cluster_logical_timestamp());
 INSERT INTO t.test(k, v, t) VALUES (3, 'fajita', cluster_logical_timestamp());
 END;