One thing that I have noticed, that you might be being bitten by, is
that if there are more than 10000 results for the select in the in
clause, then the results will be stored to disk, and not in memory. In
my testing if the in clause had 9999 elements it would return in<1s,
and if it had 10000 elements it would take about 50 seconds.
There is a parameter that you can use to increase that 10000 limit,
but I cannot recall what it was, sorry.
Well spotted.
SET MAX_MEMORY_ROWS
|SET MAX_MEMORY_ROWS int <http://www.h2database.com/html/grammar.html#int>|
The maximum number of rows in a result set that are kept in-memory. If
more rows are read, then the rows are buffered to disk. The default
value is 10000.
Admin rights are required to execute this command, as it affects all
connections. This command commits an open transaction. This setting is
persistent. It has no effect for in-memory databases.
Example:
SET MAX_MEMORY_ROWS 1000
Regards,
Peter
--
You received this message because you are subscribed to the Google Groups "H2
Database" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/h2-database?hl=en.
