Rob Klopp

Big Data Startups | The number one big data knowledge platform http://lnkd.in/hHF-ca 

@pauljoubert I agree, Paul. I'm on it... driven, if not inspired, by this twitter on the subject. Gimme a couple of weeks to do my homework.

@joeharris76 I have a connection at the NSA and I'm ripping off all of the notes and drafts you build for your blog, Joe... :)

@applebyj What changed to make the SQL faster, John? I thought that Calc views were required for performance?

@pauljoubert @vijayasankarv @timoelliott @SQLServer Not wrong. There is no in-memory column store outside of xVelocity. Check your facts.

@pauljoubert @vijayasankarv @timoelliott @SQLServer that would be xVelocity...

@pauljoubert @vijayasankarv @timoelliott @SQLServer Right, Paul. In-memory OLTP but no integrated column store to speed analytics.

@pauljoubert thanks, Paul... Will do..

@pauljoubert @vijayasankarv @timoelliott could you point us all, Paul... I worked hard on the blog post and included references...

The Big Data Devil http://wp.me/p1a7GL-ng 

@vijayasankarv @esjewett so let me be clearer here and now. HANA should perform OLTP as well as any general purpose OLTP system....

@vijayasankarv @esjewett Out of context you could argue that I glossed over the trade-off. But in-context your gripe is a little unfair.

@esjewett @ErichSap Completely agree. That is why I closed suggesting there is a place for specialized OLTP systems.

@esjewett @ErichSap It is impossible to write 1B rows/sec to any known persistent storage device... this is a rate that is all in-memory

@esjewett @ErichSap I could be wrong... but probably not disingenuous...