I have a database with many tables. This database is most part empty or has
at most 10 lines per table.
I have a query that uses many tables (~50) and uses inner/left join (always
on indexed columns (PK/FK)).
The query is really slow but not on execution, during prepareSatement.
I activated traces on the database and find out that there were 10 000 000
(at least) calls to "potential plan item cost".
There is also the same call repeatedly "calculate cost for plan [T1 ...
T30]" (where T1 ... T30 lists all the aliases of my tables.
I had to stop the database when trace file got 1Gb big.
Last cost printed in trace file was "best plan item cost
1,358,740,342,666,069,800,000,000,000,000"... It seams really big... for
such an empty database.
On other databases the query executes in less than 10ms. On H2, it needs
between 5 and 6 seconds to perform prepareStatement.
Is there a way to configure some optimization algorythm ? Or a way to not
optimize at all ?
You received this message because you are subscribed to the Google Groups "H2
To unsubscribe from this group and stop receiving emails from it, send an email
To post to this group, send email to email@example.com.
Visit this group at https://groups.google.com/group/h2-database.
For more options, visit https://groups.google.com/d/optout.