[rds-devel] netperf patch to support RDS

Or Gerlitz ogerlitz at voltaire.com
Sun Nov 4 00:09:25 PDT 2007


Richard Frank wrote:
> Or Gerlitz wrote:

>> OK, so there are actually seven numbers around:
>> UDP     - Gbe, IPoIB datagram / connected mode

> It is not valid - for performance only consideration - to compare an 
> unreliable transport / protocol to a reliable one.

fair enough.

So we are back to crload and friends, that is a tool that lets you 
compare performance over UDP with reliability done at the app level, to 
performance over RDS with reliability provided by the socket.

Such a comparison is not possible with rds-tools nor with netperf, as 
you claim here.

For that end, can we get from Oracle crload and the IPC libraries (the 
RDS and UDP based)? it seems that the only comparison possible now is 
using Oracle setup, and there must be something lighter.

> Netperf is too simplistic in that all it seems to do is stream data in a 
> simple loop. This is not how Oracle uses the IPC and again does not 
> reflect what it would take to make UDP reliable.
> For this reason we are not interested in having Netperf support RDS and 
> or seeing Netperf data.

OK

> Of course if others folks want to do this - for their own reasoning - 
> guess it's going to happen anyway.

so what's your take on pushing the RDS patch to netperf, do you think 
its wrong idea since it can mislead people?

> Just be careful to not quote Netperf data comparisons when positioning 
> an IPC for Oracle.

OK

Or.




More information about the rds-devel mailing list