tag:blogger.com,1999:blog-9149523927864751087.post2014106502614643138..comments2024-03-26T09:43:01.052-07:00Comments on Small Datum: Benchmarks vs real workloadsMark Callaghanhttp://www.blogger.com/profile/09590445221922043181noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-9149523927864751087.post-2934498755622885892020-02-15T13:41:44.687-08:002020-02-15T13:41:44.687-08:00Your project is interesting. I have run tpcc-mysql...Your project is interesting. I have run tpcc-mysql a few times but my focus has been elsewhere. I am slowly figuring out whether to use Java or golang for my next benchmark client. Mark Callaghanhttps://www.blogger.com/profile/09590445221922043181noreply@blogger.comtag:blogger.com,1999:blog-9149523927864751087.post-87101950921880506132020-02-15T12:49:38.133-08:002020-02-15T12:49:38.133-08:00Well, I never did get the opportunity to run TPC-C...Well, I never did get the opportunity to run TPC-C at a large enough scale to be able to publish the results. I have developed applications/workload that emulate TPC-C as closely as possible. TPC-C.js is an example of that. For the next implementation I intend to use either Golang or Rust so that at least the client-side of infrastructure could be used efficiently, and so it can be cheap for the vendor doing the benchmark run.<br /><br />http://gurjeet.singh.im/blog/2014/07/23/announcing-tpc-c-dot-js-a-lightweight-implementation-of-tpc-c/<br /><br />https://github.com/DBYardstick/TPC-C/tree/master/TPC-CGurjeet Singhhttps://www.blogger.com/profile/07566934368109450898noreply@blogger.comtag:blogger.com,1999:blog-9149523927864751087.post-62543073156668175772020-02-15T10:13:45.779-08:002020-02-15T10:13:45.779-08:00I assume there is a lot pressure when doing tpc-c ...I assume there is a lot pressure when doing tpc-c given that so much HW is required and that HW isn't available indefinitely.<br /><br />And then there was this story. Is that another kind of byzantine failure?<br />https://www.cnet.com/news/superdome-tests-recover-after-sabotage/Mark Callaghanhttps://www.blogger.com/profile/09590445221922043181noreply@blogger.comtag:blogger.com,1999:blog-9149523927864751087.post-63949026647132930472020-02-15T10:08:51.509-08:002020-02-15T10:08:51.509-08:00Having studied and implemented TPC-C a few times, ...Having studied and implemented TPC-C a few times, I am particularly a fan of well defined benchmarks. There's so much more involved in running a benchmark; the workload definition is a small but a critical part of it.<br /><br />I enjoyed reading your linked post on benchmarketing; it succinctly describes the enormous challenges of running benchmarks, publishihng and comparing them.Gurjeet Singhhttps://www.blogger.com/profile/07566934368109450898noreply@blogger.comtag:blogger.com,1999:blog-9149523927864751087.post-25706060484884267952020-02-14T19:46:03.713-08:002020-02-14T19:46:03.713-08:00Thank you for sharing an interesting post.
Is the...Thank you for sharing an interesting post.<br /><br />Is the star schema benchmark still in use? I used it for a while at Oracle when I worked on bitmap indexes but I am curious how well it scales (behaves with larger scale factors). Searching now I found a paper from 2009 that I will read - https://www.cs.umb.edu/~poneil/StarSchemaB.PDF<br />Mark Callaghanhttps://www.blogger.com/profile/09590445221922043181noreply@blogger.comtag:blogger.com,1999:blog-9149523927864751087.post-44553164784712577472020-02-14T16:31:16.891-08:002020-02-14T16:31:16.891-08:00The part of this problem was described in the arti...The part of this problem was described in the article: https://habr.com/en/company/yandex/blog/485096/Aleksei Milovidovhttps://www.blogger.com/profile/06488690125748102383noreply@blogger.com