Re: Lock-free databases

"Mark D Powell" <Mark.Powell_at_eds.com> wrote in message news:1131475590.080140.136430_at_g47g2000cwa.googlegroups.com...
> Technically from the theories point of view what you say is true, but I
> do not think you can create a db manager based on this concept that
> works well in a high concurrent update environment.

"ANTs 3.2 is also the only lock-free relational database management system architected for popular 64-bit Linux operating system implementations running on AMD Opteron and Intel Xeon platforms. The ANTs Data Server sets a new precedent in the database industry, allowing large OLTP, real-time analytical processing and enterprise reporting to run concurrently in the same server. "

http://www.ants.com/index.php?option=com_content&task=view&id=437&Itemid=29

http://www.eweek.com/article2/0,1895,1845154,00.asp

>
>There is going to
> be an alful lot of rework. The rework will take system resouces. The
> db manager software would probably be better off to have the first
> process wait on the in process work to complete and to apply the
> resouces to another non-waiting process in the meantime.

Performance:
http://www.cs.chalmers.se/~phs/TechnicalReports/SunT02_Noble.pdf

http://www.cs.tau.ac.il/~shanir/reading%20group/p73-michael.pdf

>
> The proof will be if someone actually develops a working product.

See above.

>
> IMHO -- Mark D Powell --
>
Received on Wed Nov 09 2005 - 01:00:28 CET