hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From mirage1987 <mirage1...@gmail.com>
Subject Re: Improving hbase performance
Date Tue, 17 Feb 2009 19:26:53 GMT

Hey thanx for the quick reply..
well i have the following problem
i have in place a distributed setup of hbase over hadoop (cluster of 3)
i have in my hbase around 4 million entries
wat i want now is an efficient way to do a get (read) from the hbase.
I have multiple read requests. So i want them to be executed in parallel so
tha the performance could imrove.(as sequential one would be too high). I
want a performance of arnd 10 ms on an average.
(Can u suggest wat is a good performance read time on a distributed hbase)

So wat i tried was creating parallel threads each with itz own HTable
instance and doin a get on the same table in parallel.
But the performance was still bad.
Could i be doin some mistake.....is it possible to create multiple instances
of hbase that could be accessed in parallel.

Second thought i have is can i do the same thing with Mapreduce...
Can u suggest how i could do this.....some examples i have seen suggest that
they are just scanning through the whole database which would perform even
worse in my case.
please help me on this

i still dont understand how the hbase is working...
-- 
View this message in context: http://www.nabble.com/Improving-hbase-performance-tp22060928p22064387.html
Sent from the HBase User mailing list archive at Nabble.com.


Mime
View raw message