In particular, TPCC requires transactions to be compliant, and we don't have transactions, so it hasn't been a goal.
We should probably re-run the YCSB benchmarks at some point soon since there have been lots of performance optimizations since 0.5, and I imagine many of them would improve the results. -Todd On Wed, Sep 14, 2016 at 4:20 PM, Roberta Marton <roberta.mar...@esgyn.com> wrote: > Thanks > > > > Roberta > > > > *From:* David Alves [mailto:davidral...@gmail.com] > *Sent:* Wednesday, September 14, 2016 4:17 PM > *To:* user@kudu.apache.org > *Subject:* Re: Question on performance of Kudu > > > > Hi Roberta > > > > We have some (old) ycsb numbers vs hbase on the kudu technical > whitepaper: https://kudu.apache.org/kudu.pdf > > We haven't ran any benchmarks recently though, and we never tried tpcc > iirc. > > > > Best > > David > > > > On Wed, Sep 14, 2016 at 4:12 PM, Roberta Marton <roberta.mar...@esgyn.com> > wrote: > > Have you run any comparisons with Kudu using common benchmarks such as > TPCC to see how Kudu compares against HBase and others? > > With your innovative design, these tests should show some good numbers. > > > > Regards, > > Roberta Marton > > > -- Todd Lipcon Software Engineer, Cloudera
