Agneeswaran,

I have tested and profiled ODBC using data from odbc-example and I can see
that 1 million rows, 3 string columns each retrieved by ODBC in ~8 seconds
on
first run, and in less than ~5 seconds on second run.

Actually, ODBC driver always retrieving data by pages of size 1024 rows so
it
should be fast for the most use-cases.

Please, check your configuration. In particular, if you've used
odbc-example as
a template, please, check that you have set proper value for the
cfg.jvmMaxMem as the value in the example is low for a big number of records
which can cause a lot of GC.


Best Regards,
Igor

On Wed, Jun 29, 2016 at 5:16 PM, Igor Sapego <isap...@gridgain.com> wrote:

> Agneeswaran,
>
> Let me have a look. I will profile the code and will tell you results.
>
> Best Regards,
> Igor
>
> On Wed, Jun 29, 2016 at 4:48 PM, Agneeswaran <
> agneeswaran.ponnuraman...@nielsen.com> wrote:
>
>> Hi Igor,
>>
>> Using Java API we can fetch the data within ~5 seconds but in C++ it took
>> ~4
>> minutes.
>>
>> Will it possible to fetch the data, let say  500-1000 records at once.
>>
>> Could you please suggest us.
>>
>> Thanks,
>> Agneeswaran
>>
>>
>>
>> --
>> View this message in context:
>> http://apache-ignite-users.70518.x6.nabble.com/Performance-issue-on-Ignite-ODBC-API-tp5908p5998.html
>> Sent from the Apache Ignite Users mailing list archive at Nabble.com.
>>
>
>

Reply via email to