Re: [Lsr] Doubt regarding A bit set/clear

["Les Ginsberg (ginsberg)" <[e-mail]>]

Gurusiddesh –

The short answer to all your questions is “yes”.
More inline.

From: Gurusiddesh Nidasesi <[e-mail]>
Sent: Thursday, April 22, 2021 10:33 PM
To: Acee Lindem (acee) <[e-mail]>
Cc: [e-mail]; [e-mail]; [e-mail]; Les Ginsberg (ginsberg) <[e-mail]>; Dona Maria John <[e-mail]>; Vikram Agrawal <[e-mail]>; Mahalakshmi Kumar <[e-mail]>
Subject: Re: [Lsr] Doubt regarding A bit set/clear

Hi All.

Gentle Reminder!

Regards,
Gurusiddesh V N

On Mon, Apr 19, 2021 at 6:54 PM Acee Lindem (acee) <[e-mail]<mailto:[e-mail]>> wrote:
Gurusiddesh,

I’ll defer to the RFC authors on your question. However, please refrain from referring to bits as being “unset”. They are set or clear.

Thanks,
Acee

From: Lsr <[e-mail]<mailto:[e-mail]>> on behalf of Gurusiddesh Nidasesi <[e-mail]<mailto:[e-mail]>>
Date: Monday, April 19, 2021 at 6:13 AM
To: "[e-mail]<mailto:[e-mail]>" <[e-mail]<mailto:[e-mail]>>
Cc: Spencer Giacalone <[e-mail]<mailto:[e-mail]>>, Stefano Previdi <[e-mail]<mailto:[e-mail]>>, "Les Ginsberg (ginsberg)" <[e-mail]<mailto:[e-mail]>>, Dona Maria John <[e-mail]<mailto:[e-mail]>>, Vikram Agrawal <[e-mail]<mailto:[e-mail]>>, Mahalakshmi Kumar <[e-mail]<mailto:[e-mail]>>
Subject: [Lsr] Doubt regarding A bit set/unset

Hi All,

I had a query regarding setting/unsetting A bit.

https://tools.ietf.org/html/rfc8570#section-4.1 states that

A bit:  This field represents the Anomalous (A) bit.  The A bit is

      set when the measured value of this parameter exceeds its

      configured maximum threshold.  The A bit is cleared when the

      measured value falls below its configured reuse threshold.  If the

      A bit is cleared, the sub-TLV represents steady-state link

      performance.



So does it mean we have to have two configurations one for reuse and another for maximum threshold?

[Les:] Yes. The goal is to prevent altering the advertisement due to small oscillations of the advertisement. If you had a single value then if the measured value bounced between (for example) +1/-1 of the threshold) the advertisement of the A-bit would change rapidly – this is undesirable.

So the max threshold triggers setting of the A-bit and the reuse threshold triggers clearing of the bit. The reuse threshold provides some confidence that the measurement has stabilized below the maximum anomalous threshold.





Will below example follow the RFC?

reuse threshold : 50 usec

maximum threshold : 100 usec

1st measured value : 110 usec

conclusion: Set A bit.



2nd measured value : 75 usec

conclusion : Do nothing (Maintain pervious state of A bit as the value is less than reuse or greater than threshold)



3rd measurement value : 30 usec

conclusion: Unset A bit.



[Les:] Yes this conforms to specified behavior.



 If we have to have two configuration for threshold to set/unset A bit, will they be different from the threshold that we use for advertisements?

[Les:] Yes .  This is clearly stated in Section 5:



“4.  For sub-TLVs that include an A bit, an additional threshold

       SHOULD be included corresponding to the threshold for which the

       performance is considered anomalous (and sub-TLVs with the A bit

       are sent)…”
   Les



Regards,
Gurusiddesh V N

.


--
Thanks,
Gurusiddesh V N

.