Hi John,
Thanks for the reply,
We are still working on the sample data available with fastbit
examples. To reach 5G partitions, we are concatenating the data
available in test0.csv under tests directory of Fastbit. Once we
achieve the desired performance with fastbit we will start
experimenting Fastbit with our business data.
Kindly find below answers to your questions:
1. How much memory do you have on the test machine? how much memory is
allotted to FastBit?
We are running the fastbit on a dedicated Redhat linux el5 machine
without much overhead from other processes. Test machines have a
memory of 24G, with around 8G allotted to Fastbit . (This is set
through java argument)
2. How many results (rows) are you retrieving? what type of values
are involved? what is the expected sizes of them?
Around 4497792 rows out of total <> rows are being
retrieved per partition. We are fetching columns with integers types.
Each query results in 5MB data.
3. Are you keeping the values in Java code? How much memory do you
expect to use?
We are not keeping the values in java code. Just to analyze this
issue in detail, we are throwing away the result. Once we get pass
this issue, we will start utilizing the results values further in our
code.
Many Thanks,
Ghousia.
_______________________________________________
FastBit-users mailing list
[email protected]
https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users
