KB2499023-Labojums: nepareizi rezultāti, ja komplekss atjaunināšanas vaicājums izmanto SAVIENOJUMA operatoru izpildes plānā programmā SQL Server 2008 vai SQL Server 2008 R2

Simptomi

Ja palaižat sarežģītu atjaunināšanas vaicājumu, kas izmanto savienojuma operatoru izpildes plānā programmā Microsoft sql Server 2008 vai Microsoft sql Server 2008 R2, var tikt parādīti nepareizi rezultāti. Piezīme. Atjaunināšanas vaicājumā var būt Apakšvaicājumi, Uzkrātās funkcijas un atšķirīga klauzula. Ja to pašu vaicājumu izpildāt programmā Microsoft SQL Server 2008 vai Microsoft SQL Server 2008 R2 paralēlās datu noliktavā, iespējams, saņemsit citu nepareizo rezultātu kopu. Piemēram, palaižot šādu kodu, var tikt parādīti nepareizi rezultāti:

UPDATE RB SET rb1 = (SELECT AVG(DISTINCT T2_1.C1) AS C1 FROM (SELECT DISTINCT T3_1.C3 AS C1, T3_1.C5 AS C2 FROM (SELECT T5_1.c2 AS C1, T5_1.c2 AS C2, T5_1.c3 AS C3, T5_1.c3 AS C4, T5_1.c2 AS C5 FROM DC AS T5_1 UNION SELECT T5_1.c2 AS C1, T5_1.c3 AS C2, T5_1.c2 AS C3, T5_1.c2 AS C4, T5_1.c3 AS C5 FROM DC AS T5_1) AS T3_1) AS T2_1 WHERE (T2_1.C1 = rb3)), rb2 = (SELECT COUNT_BIG(DISTINCT T2_1.C1) AS C1 FROM (SELECT DISTINCT T3_1.C1 AS C1 FROM (SELECT T5_1.rc2 AS C1 FROM RC AS T5_1 UNION ALL SELECT T5_1.c3 AS C1 FROM DC AS T5_1) AS T3_1) AS T2_1 WHERE (T2_1.C1 > rb1)), rb3 = (SELECT COUNT_BIG(*) AS C1 FROM (SELECT DISTINCT T3_1.C1 AS C1, T3_1.C2 AS C2, T3_1.C3 AS C3, T3_1.C4 AS C4, T3_1.C5 AS C5, T3_1.C6 AS C6, T3_1.C7 AS C7, T3_1.C8 AS C8, T3_1.C9 AS C9, T3_1.C10 AS C10 FROM (SELECT T5_1.b2 AS C1, T5_1.b1 AS C2, T5_1.b3 AS C3, T5_1.b2 AS C4, T5_1.b2 AS C5, T5_1.b2 AS C6, T5_1.b2 AS C7, T5_1.b1 AS C8, T5_1.b2 AS C9, T5_1.b3 AS C10 FROM DB AS T5_1 UNION SELECT T5_1.ra3 AS C1, T5_1.ra3 AS C2, T5_1.ra3 AS C3, T5_1.ra3 AS C4, T5_1.ra2 AS C5, T5_1.ra3 AS C6, T5_1.ra1 AS C7, T5_1.ra1 AS C8, T5_1.ra1 AS C9, T5_1.ra1 AS C10 FROM RA AS T5_1) AS T3_1) AS T2_1 WHERE (T2_1.C2 <= rb3)) WHERE rb1 IN (SELECT T2_1.C1 AS C1 FROM (SELECT T4_1.a2 AS C1, T4_1.a1 AS C2, T4_1.a1 AS C3, T4_1.a2 AS C4, T4_1.a1 AS C5, T4_1.a2 AS C6, T4_1.a3 AS C7, T4_1.a1 AS C8 FROM DA AS T4_1 WHERE (T4_1.a1 > (SELECT MIN(DISTINCT T5_1.b3) AS C1 FROM DB AS T5_1 WHERE (T5_1.b3 <= T4_1.a1))) UNION ALL SELECT T4_1.a2 AS C1, T4_1.a2 AS C2, T4_1.a1 AS C3, T4_1.a1 AS C4, T4_1.a1 AS C5, T4_1.a1 AS C6, T4_1.a1 AS C7, T4_1.a3 AS C8 FROM DA AS T4_1) AS T2_1);

Cēlonis

Šī problēma rodas kļūdas dēļ SQL Server vaicājumu optimizētājs.

Risinājums

Statusa

Microsoft ir apstiprinājusi, ka šī problēma pastāv Microsoft produktos, kas ir norādīti sadaļā "attiecas uz".

Atsauces

Lai iegūtu papildinformāciju par paralēlo datu noliktavu, apmeklējiet šo Microsoft tīmekļa vietni:

Vispārīga informācija par paralēlo datu noliktavuLai iegūtu papildinformāciju par papildu apkalpošanas modeli SQL Server, noklikšķiniet uz šī raksta numura, lai skatītu Microsoft zināšanu bāzes rakstu:

935897 SQL Server komandā ir pieejams papildu apkalpošanas modelis, kas sniedz labojumfailus ziņotām problēmāmLai iegūtu papildinformāciju par nosaukumdošanas shēmu SQL Server atjauninājumiem, noklikšķiniet uz šī raksta numura, lai skatītu Microsoft zināšanu bāzes rakstu:

822499 Jauna nosaukuma shēma Microsoft SQL Server programmatūras atjaunināšanas pakotnēmLai iegūtu papildinformāciju par programmatūras atjaunināšanas terminoloģiju, noklikšķiniet uz šī raksta numura, lai skatītu Microsoft zināšanu bāzes rakstu:

824684 Apraksts par standarta terminoloģiju, kas tiek izmantota Microsoft programmatūras atjauninājumu aprakstos