[ofa-general] some comments/clarifications on ipoib/opensm re qos

[Yevgeny Kliteynik]

Or Gerlitz wrote:
> Yevgeny Kliteynik wrote:
>> OK. Can you rephrase what you said in two sentences? Something short
>> to replace the existing (and wrong, as you pointed out) three lines:
>>
>>   151 IPoIB queries the SA for its broadcast group information.
>>   152 It provides the broadcast group SL, MTU, and RATE in every 
>> following
>>   153 PathRecord query performed when a new UDAV is needed by IPoIB.
> Will try to send something later this week.

Thanks

> This document describes the QoS related architecture of Linux IB 
> stack, hence do you agree that the file name would be changed to 
> reflect that? ie that the word "ofed" will not appear there?

Not sure I'm following.
This doc describes OFED components (SDP, RDS, SRP, IPoIB, iSER),
their support of QoS annex, and a way for a QoS manager (OpenSM,
which is part of OFED) to enforce the QoS policy.
I really don't have any special sentiments for file name, but
is QoS_in_OFED.txt wrong?
You probably meant something like QoS_architecture.txt?

>>> This creates confusion and I am not sure its well defined, at least 
>>> in my case when I put my hands on QoS testing, I couldn't guess that 
>>> "ipoib" == "pkey 0x7fff". For example, is it correct that "ipoib, 
>>> pkey 0x8001 : <SL>" is not valid rule?
>>
>> This rule is also valid, providing that you actually have a partition
>> with pkey 0x8001 configured in the partition cfg file.
> OK, so the rule is valid, but would you agree that the word "ipoib" 
> adds nothing to the rule?

Agree, it's just an alias.
 'ipoib : <SL>' is useful
 'any, pkey 0xNNNN' is useful too.
 'ipoib, pkey 0xNNNN' is just for better readability of the matching rules.

-- Yevgeny

>
> Or.
>
>
> _______________________________________________
> general mailing list
> general at lists.openfabrics.org
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
>
> To unsubscribe, please visit 
> http://openib.org/mailman/listinfo/openib-general
>