[cgl_discussion] CGL 3.0 performance draft announcement

Eric.Chacron at alcatel.fr Eric.Chacron at alcatel.fr
Fri May 7 02:07:12 PDT 2004

Adriano, Peter,

I'd like to add that we are today carefull with emergent technologies like
even if it looks promising ( just because we dont use it and maybe dont
plan to use it at that time
whereas gibabit Ethernet is already used ).

About RDMA,  the spec already includes a general requirement for low
software latency ( PRF.33.1)
"Zero-copy based on memory mapping and simplified path from user to
physical layer
( reducing abstract layers overhead) shall be used."
I have listed VIA in the note section, and i think VIA architecture uses

About iSCSI and SAN i think we could think more about storage performance,
to specify some requirement
that we have maybe missed in the first draft. I agree with you Peter that
this is an area related to
hardware and cluster too.


Peter Badovinatz <tabmowzo at us.ibm.com>@lists.osdl.org on 05/06/2004
07:34:58 PM

Sent by:    cgl_discussion-bounces at lists.osdl.org

To:    Adriano Galano <adriano at satec.es>
cc:    Eric CHACRON/FR/ALCATEL at ALCATEL, cgl_discussion at osdl.org
Subject:    Re: [cgl_discussion] CGL 3.0 performance draft announcement

Adriano Galano wrote:

> Hello:
> Nothing about Infiniband, iSCSI,  RDMA, clusters or grid filesystems in
> draft?


We have a separate spec draft dedicated to clustering (and including
cluster file/data access), see:

I can't say that we've spent any real time discussing grids, most of
what we anticipate as carrier applications we will support will be in
clustered environments, but not really in grids (I define grid here as
the 'utility' computing view), since the systems are more 'fixed'.

Infiniband, iSCSI and RDMA have, likewise, not been items brought up.
We'd likely view them as methods to get function or performance, so we
would tend not to prescribe them as how you have to build the system but
they could be optional components.  We are working on a Hardware
specification, and a discussion of these could take place for that.
Aspects of these would fall into our clustering view as well.

> Regards,
> -Adriano
> -----Mensaje original-----
> De: cgl_discussion-bounces at lists.osdl.org
> [mailto:cgl_discussion-bounces at lists.osdl.org] En nombre de
> Eric.Chacron at alcatel.fr
> Enviado el: jueves, 06 de mayo de 2004 14:35
> Para: cgl_discussion at osdl.org
> Asunto: [cgl_discussion] CGL 3.0 performance draft announcement
> Hi,
> I would like to announce the availability of an
> early public draft of the Carrier Grade Linux v3.0 Performance
> specification, available at:
> http://www.osdl.org/docs/cgl_perf_req_def_30.pdf
> This draft includes requirements made from telecom industry perspective
> on Linux performances and scalability.
> Thanks in advance for your comments.
> -Eric
> Notes:
> -I acknowledge that the requirements in this draft are being implemented
> in a variety of ways and many of the requirements in this document exist
> in current implementations.  I am contacting this mailing list because I
> believe your projects and expertise may address some of the requirements
> and we'd like to solicit feedback.
> -Again, this is an early draft document of the v3.0 performance
> requirements spec.  Past OSDL Carrier Grade Linux technical documents
> have contained all requirements in a single document.  For OSDL CGL
> v3.0 draft releases, we are releasing them as more granular sections,
> roughly split on functional boundaries.  These boundaries are
> Standards, Availability, Clustering, Hardware,
> Performance (this document), Security, and Serviceability .
> -More information on Carrier Grade Linux and the Carrier Grade Linux
> Working Group can be found at:
> http://osdl.org/lab_activities/carrier_grade_linux/.
> -Feel free to direct any comments on the spec to me directly at
> eric.chacron at alcatel.fr or to cgl_discussion at osdl.org.

Peter R. Badovinatz aka 'Wombat' -- IBM Linux Technology Center
preferred: tabmowzo at us.ibm.com / alternate: wombat at us.ibm.com
These are my opinions and absolutely not official opinions of IBM, Corp.

cgl_discussion mailing list
cgl_discussion at lists.osdl.org

More information about the cgl_discussion mailing list