Seems like a result of Alverro's cluster patch -- looks like the patch
didn't updated the expected results for the regression tests
fully. Diffs below.
Cheers,
Neil
--
Neil Conway <neilc@samurai.com> || PGP Key ID: DB3C29FC
*** ./expected/cluster.out Fri Nov 15 12:35:36 2002
--- ./results/cluster.out Fri Nov 15 12:39:33 2002
***************
*** 302,307 ****
--- 302,310 ---- INSERT INTO clstr_2 VALUES (1); INSERT INTO clstr_3 VALUES (2); INSERT INTO clstr_3 VALUES (1);
+ -- "CLUSTER <tablename>" on a table that hasn't been clustered
+ CLUSTER clstr_2;
+ ERROR: CLUSTER: No previously clustered index found on table clstr_2 CLUSTER clstr_1_pkey ON clstr_1; CLUSTER
clstr_2_pkeyON clstr_2; SELECT * FROM clstr_1 UNION ALL
***************
*** 344,349 ****
--- 347,364 ---- 1 (6 rows)
+ -- cluster a single table using the indisclustered bit previously set
+ DELETE FROM clstr_1;
+ INSERT INTO clstr_1 VALUES (2);
+ INSERT INTO clstr_1 VALUES (1);
+ CLUSTER clstr_1;
+ SELECT * FROM clstr_1;
+ a
+ ---
+ 1
+ 2
+ (2 rows)
+ -- clean up \c - DROP TABLE clstr_1;