|
 STATISTICS in MS SQL Server are objects
somewhat similar to indexes and occupying the same namespace. They
help the server to optimize queries. Statistics may be created
either explicitly or by the server itself. In the Diff, you
can synchronize statistics or ignore them, as described below:
The "Don't load" setting means the Diff
will totally ignore any STATISTICS, so it won't even be able to
resolve dependencies if a particular STATISTICS object blocks
changes to the table. We recommend that you don't use this
setting until you know exactly what you are doing.
With the "Ignore changes" option, the explicitly
specified STATISTICS will be loaded and compared, but any changes
in them will not propagate up the schema tree. So if a pair of
tables only differs in STATISTICS, it will be marked as unchanged
and the Diff won't try to synchronize them. When there are changes
in columns, the Diff will take the statistics into account,
removing them when necessary and re-creating afterwards. Note
that it will try to "retain" statistics, that is re-create them as
they were in the target
database, not in the source one. You still can synchronize
STATISTICS if you specifically select any of them of the entire
"Statistics" collection in the schema tree.
The "Compare and synchronize" option tells
the Diff to treat STATISTICS as any other schema object to be
compared and synchronized.
Note that regardless of the
settings above, the Diff never loads any statistics
automatically created by the server
The "Ignore names..." checkboxes tells the
Diff that STATISTICS or INDEXes defined on the same combination of
columns are equivalent regardless of their names.
|
