Redfish on OpenBMC

[Yugi Mani]

On Thursday, February 8, 2018 6:17 AM, Rapkiewicz, Pawel [pawel.rapkiewicz at intel.com] wrote:

> - gevent and uwsgi, those are two independent redfish server
> implementations?
Yes, we had gevent and uwsgi on AST2400. 
gevent and cherrypy on AST2500. 
Chris's results are using cherrypy. 
We are not supporting gevent anymore on both platforms because of its lack of concurrency ability. 

> - do you have any traffic during retrieving root service?
System Under Test (SUT) was on corporate network. System might have received some broadcast packets, but shouldn't have received any unicast packets except for Redfish commands. 

> - Was it with Basic HTTP Auth, or without any?
As Hari identified before service root requests were unauthenticated and other requests that we benchmarked were with basic auth. 

> Code size:
> - real code size, not on compressed file system
We set this flag "BB_GENERATE_MIRROR_TARBALLS" in our conf file. Size I sent was for tar file. 

> - what's the size of dependent libraries (already existing in OpenBMC
> environment vs. the new one required only by given Redfish
> implementation)?
I will calculate and let you know. 

 > Performance:
> - We need common script for measurement
If you are using curl on linux, them "time curl ..." would do. 

> - And common method, I'd prefer running on machine directly
> connected to the OpenBMC platform to minimize network latency, or to
> run it over loopback interface on the platform, but this may affect
> context switching
We did it on regular corporate network. 

> - The performance will still be biased by the CPU the Redfish is running,
> but it will be closer to make it comparable.
There are lots of parameters that can affect Redfish response time - number of processes running, priority of redfish service, size of response, how redfish gathers response data, what is the format of source data, apart from CPU & Network that you identified. 
It is really hard to do apples-apples as you said.