Panayiotis Vlissidis,

  Thanks a lot for the reply. I made the changes as you suggested.
Created multi-column indexes.

   The time taken to update the table is 1.5 sec instead of 2 sec with
MULTI_THREADED=1.


Thanks,
Suresh






On Feb 14, 6:36 pm, suresh <[email protected]> wrote:
> Hi,
>
>        I am using in-memory h2 db and I have 8 tables. Each table has
> 15 to 20 columns. One table with 60 columns.  I have two indexes per
> table. Each table will have at most 3k rows.
>
>        Every 5 sec I am updating these tables with new data. I am
> updating/inserting/deleting based on the delta.
> It is real-time data feed. At the same time there will be 100 read
> requests on these tables.
>
> I am not using any transactions. Plain update/select queries.
>
> I have tried with following two options :
>
> 1) MVCC=TRUE
>
>    The time took to update a table is 4 sec.
>    The read queries takes 100 ms for each query
>
> 2) MULTI_THREADED=1
>    The time took to update the a table is 2 sec.
>    The read queries takes 100 ms for each query
>
> I am using h2-1.2.147 version of h2. I would like to know what is the
> best way to update the rows in the table less than 1 sec.
> Each update contains less than 500 rows per table.
>
> I am finding interesting thing in TRACE options.
>
> /*SQL #:1 t:200*/UPDATE CURRENT_VALUES SET ATT1 = 9, ATT2 = 141, ATT3=
> '141', \n\t\tATT4 = 3745, ATT5 = '3745', ATT5 = 232, ATT6 = '232',
> ATT7 = 0, \n\t\tATT8 = '', ATT9 = 'xyz', ATT10 = 1297689071078, AT11 =
> 0, ATT12 = 0,\n\t\tATT13 = 101 \n\t\tWHERE ATT2 = 141 AND ATT3 = 3745
> AND ATT4 = 232 AND ATT5 = 0 AND ATT9 = 'xyz';
>
> 02-14 13:11:47 command: slow query: 1,091 ms
>
> The above query took more than 1 sec per row update.
>
> There is select query it took  1,429 ms, the table has 60 columns.
>
> Is there any way to optimize the performance?
>
> Thanks,
> Suresh

-- 
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.

Reply via email to