diff options
| author | Tom Lane <tgl@sss.pgh.pa.us> | 2008-12-13 19:13:44 +0000 |
|---|---|---|
| committer | Tom Lane <tgl@sss.pgh.pa.us> | 2008-12-13 19:13:44 +0000 |
| commit | 65e3ea76417d1baab158fd8305ebed4f43141c7a (patch) | |
| tree | 289f332ba577bcbd9794c3e11ef2b3b89cb078be /src/backend/commands/analyze.c | |
| parent | b69bde774982ac4497cec9c9fe0190097890292f (diff) | |
| download | postgresql-65e3ea76417d1baab158fd8305ebed4f43141c7a.tar.gz postgresql-65e3ea76417d1baab158fd8305ebed4f43141c7a.zip | |
Increase the default value of default_statistics_target from 10 to 100,
and its maximum value from 1000 to 10000. ALTER TABLE SET STATISTICS
similarly now allows a value up to 10000. Per discussion.
Diffstat (limited to 'src/backend/commands/analyze.c')
| -rw-r--r-- | src/backend/commands/analyze.c | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/src/backend/commands/analyze.c b/src/backend/commands/analyze.c index 6b95075be1f..2b3af54ff05 100644 --- a/src/backend/commands/analyze.c +++ b/src/backend/commands/analyze.c @@ -8,7 +8,7 @@ * * * IDENTIFICATION - * $PostgreSQL: pgsql/src/backend/commands/analyze.c,v 1.128 2008/11/10 00:49:37 tgl Exp $ + * $PostgreSQL: pgsql/src/backend/commands/analyze.c,v 1.129 2008/12/13 19:13:44 tgl Exp $ * *------------------------------------------------------------------------- */ @@ -67,7 +67,7 @@ typedef struct AnlIndexData /* Default statistics target (GUC parameter) */ -int default_statistics_target = 10; +int default_statistics_target = 100; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -1531,10 +1531,10 @@ std_typanalyze(VacAttrStats *stats) * error in bin size f, and error probability gamma, the minimum * random sample size is * r = 4 * k * ln(2*n/gamma) / f^2 - * Taking f = 0.5, gamma = 0.01, n = 1 million rows, we obtain + * Taking f = 0.5, gamma = 0.01, n = 10^6 rows, we obtain * r = 305.82 * k * Note that because of the log function, the dependence on n is - * quite weak; even at n = 1 billion, a 300*k sample gives <= 0.59 + * quite weak; even at n = 10^12, a 300*k sample gives <= 0.66 * bin size error with probability 0.99. So there's no real need to * scale for n, which is a good thing because we don't necessarily * know it at this point. |