Thanks for the result, Mateusz! I was wondering if you ran clean-database.sh between the runs and how full is the disk partition. Just want to make sure we're comparing apple with apple here.
__Luke On Feb 13, 4:20 pm, Mateusz Berezecki <[email protected]> wrote: > Here are the results: > > First, Intel C++ compiler 11 > > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_write_test 10000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 0.22 s > Total inserts: 10000 > Throughput: 46265606.64 bytes/s > Throughput: 45717.00 inserts/s > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_read_test 10000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 2.31 s > Total scanned: 10000 > Throughput: 4381625.75 bytes/s > Throughput: 4329.67 scanned cells/s > > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_write_test 1000000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 94.27 s > Total inserts: 1000000 > Throughput: 10735123.89 bytes/s > Throughput: 10607.83 inserts/s > > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_read_test 1000000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 516.12 s > Total scanned: 1000000 > Throughput: 1960775.13 bytes/s > Throughput: 1937.52 scanned cells/s > > Next is G++ 4.3.2 > > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_write_test 10000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 0.17 s > Total inserts: 10000 > Throughput: 58156908.72 bytes/s > Throughput: 57467.30 inserts/s > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_read_test 10000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 2.34 s > Total scanned: 10000 > Throughput: 4319800.09 bytes/s > Throughput: 4268.58 scanned cells/s > > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_write_test 1000000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 98.46 s > Total inserts: 1000000 > Throughput: 10278464.00 bytes/s > Throughput: 10156.58 inserts/s > r...@monkey:/opt/hypertable/0.9.2.1/bin# ./random_read_test 1000000000 > > 0% 10 20 30 40 50 60 70 80 90 100% > |----|----|----|----|----|----|----|----|----|----| > *************************************************** > Elapsed time: 552.88 s > Total scanned: 1000000 > Throughput: 1830425.10 bytes/s > Throughput: 1808.72 scanned cells/s > > It's clear that G++ 4.3.2 performs better on a smaller data set, while > intel outperforms g++ on random read test by 36.76 seconds. > > The test was run in a single node local fs environment. > Machine spec: > > 4GB non ECC RAM, > Extended brand string: "Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz" > > single SATA II disk was used during the test (no RAID) > > Mateusz --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Hypertable Development" 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/hypertable-dev?hl=en -~----------~----~----~----~------~----~------~--~---
