Mercea Paul wrote:

>
>...
>
> and make triggers on interested table like:
>
> CREATE OR ALTER TRIGGER mytable_COUNTROWS FOR mytable
> ACTIVE AFTER INSERT POSITION 0
> AS
> begin
> if (not exists(select 1 from tablesrows t where t.table_name='mytable')
> then
> insert into tablesrows (table_name, table_rows) values ('mytable',1);
>
> update tablesrows set table_rows = (select count(id) from mytable);
> end
> ^

That's a great way to induce update conflicts - every transaction
that adds, changes, or removes rows from the table has to update a
single gatekeeper record.

If you don't care about how accurate your count is, you can guess
at it from the number of RDB$PAGES entries there are for pointer
pages for the table.

First, you'll have to find the RDB$RELATION_ID for the table -

select rdb$relation_id from rdb$relations
where rdb$relation_name = UPPER (<tablename>);

Then, use gstat to get an estimate of the number of records per
data page.

gstat -a >gstat.txt

Then, compute the approximate number of data pages per pointer
page

dp_per_pp =~ <page size> / 8

Then get the number of pointer pages in the table

select count(*) from rdb$pages
where rdb$page_type = 4 and rdb$relation_id = <saved id>


Multiply the count by the number of records per page and the
number of data pages per pointer page.

Don't use rdb$db_key - there are huge gaps in the number space.


Regards,


Ann