Skip to Content.
Sympa Menu

perfsonar-dev - Re: [pS-dev] Data units stored in MAs

Subject: perfsonar development work

List archive

Re: [pS-dev] Data units stored in MAs


Chronological Thread 
  • From: Martin Swany <>
  • To: Nicolas Simar <>
  • Cc: "" <>
  • Subject: Re: [pS-dev] Data units stored in MAs
  • Date: Thu, 24 May 2007 10:15:02 +0200

Hi all,

I think that there are two ways to look at this question: from the
perspective of the perfSONAR $X.$Y suite (a given bit of software)
and from the protocol level.

What happens if one RRD MA instance export bps, another one kBps?
1) Do the visualisation tool have to deal with it?

In my opinion, the client should have to deal with the units.
From the protocol perspective, we should make sure that
1)there is a way to express the units
2)that expression is done in a standard way (as Jeff pointed
out.) Although note that 'b' vs 'bit' is still being debated.

2) Do the MA have to normalise it?

I agree with Roman that the MA should not be required to
normalize it. I also agree that this a transformation function,
but I think something this small could very well be a simple
filter that a give MA may load or not. I also think that the
transformation should be expressed the same way whether
this functionality is in the same service or a separate one
(i.e. with a transformation/operation metadata).

3) Are there best practices we should suggest to ensure consistency in the service offered globally?

From the protocol perspective, we need only to eliminate ambiguity.
From the software perspective, there are certainly "best practices",
but I would say that trying to *ensure* consistency is less desirable
than ensuring lack of ambiguity.

This also points to a way to discover what transformations a given
service will offer. That should probably be service-level metadata.

best,
martin






Archive powered by MHonArc 2.6.16.

Top of Page