percona/percona-toolkit
1
0
Fork 0
mirror of https://github.com/percona/percona-toolkit.git synced 2025-09-10 13:11:32 +00:00
Code Issues Packages Projects Releases Wiki Activity

Don't get min/max a 2nd time in _chunk_char(). Use where in _chunk_char(). Test char chunking world_city.name with real range stats. Add failing pt-table-checksum char chunk --where test.

Browse Source
This commit is contained in:
Daniel Nichter
2011-08-27 11:17:28 -06:00
parent 2e0f607589
commit a7ab27bb54
5 changed files with 105 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
lib/TableChunker.pm
View File

@@ -204,6 +204,7 @@ sub find_chunk_columns {
# exact - Use exact chunk_size? Use approximates is not.
# tries - Fetch up to this many rows to find a non-zero value
# chunk_range - Make chunk range open (default) or openclosed
# where - WHERE clause.
#
# Returns:
# Array of WHERE predicates like "`col` >= '10' AND `col` < '20'",
@@ -510,6 +511,9 @@ sub _chunk_numeric {
# <TableChunker::get_range_statistics()>
# chunk_size - requested size of each chunk
#
# Optional Arguments:
# where - WHERE clause.
#
# Returns:
# Array of chunker info that <calculate_chunks()> uses to create
# chunks, like:
@@ -522,7 +526,7 @@ sub _chunk_numeric {
# (end code)
sub _chunk_char {
my ( $self, %args ) = @_;
my @required_args = qw(dbh db tbl tbl_struct chunk_col rows_in_range chunk_size);
my @required_args = qw(dbh db tbl tbl_struct chunk_col min max rows_in_range chunk_size);
foreach my $arg ( @required_args ) {
die "I need a $arg argument" unless defined $args{$arg};
}
@@ -533,15 +537,8 @@ sub _chunk_char {
my $row;
my $sql;
# Get what MySQL says are the min and max column values.
# For example, is 'a' or 'A' the min according to MySQL?
$sql = "SELECT MIN($chunk_col), MAX($chunk_col) FROM $db_tbl "
. "ORDER BY `$chunk_col`";
MKDEBUG && _d($dbh, $sql);
$row = $dbh->selectrow_arrayref($sql);
my ($min_col, $max_col) = ($row->[0], $row->[1]);
# Get the character codes between the min and max column values.
my ($min_col, $max_col) = @{args}{qw(min max)};
$sql = "SELECT ORD(?) AS min_col_ord, ORD(?) AS max_col_ord";
MKDEBUG && _d($dbh, $sql);
my $ord_sth = $dbh->prepare($sql); # avoid quoting issues
@@ -642,7 +639,9 @@ sub _chunk_char {
# [ant, apple, azur, boy]. We assume data is more evenly distributed
# than not so we use the minimum number of characters to express a chunk
# size.
$sql = "SELECT MAX(LENGTH($chunk_col)) FROM $db_tbl ORDER BY `$chunk_col`";
$sql = "SELECT MAX(LENGTH($chunk_col)) FROM $db_tbl "
. ($args{where} ? "WHERE $args{where} " : "")
. "ORDER BY `$chunk_col`";
MKDEBUG && _d($dbh, $sql);
$row = $dbh->selectrow_arrayref($sql);
my $max_col_len = $row->[0];