Re: [j-nsp] Best way to do QOS bleach

2016-10-17 Thread Aaron
"bleach" is a new term for me, I'm assuming you mean re-marking to zero, correct? (sounds like something I would do at my qos cloud edge, aka qos trust boundary) If I use ACX5048's, which I don't think are trio-chips, but rather, are broadcomm, what method is there available for "bleaching" on AC

Re: [j-nsp] Best way to do QOS bleach

2016-10-17 Thread Olivier Benghozi
In 14.2R3 and later, and in 15.1F and 16.1R (but not in 15.1R). > On 17 oct. 2016 at 18:11, Dragan Jovicic wrote : > > And if you require more granular ingress remark, as Mark suggested after > 14.2R3.8 you can use policy-maps. ___ juniper-nsp mailing

Re: [j-nsp] Best way to do QOS bleach

2016-10-17 Thread Dragan Jovicic
If you just want to do a sort of default classifier you can use #set class-of-service interfaces all unit * forwarding-class be which at least gets rid of NC queue issue you mentioned. If you want to actually bleach customer dscp on ingress, you can do it with MF classifier. There are several dr

Re: [j-nsp] Best way to do QOS bleach

2016-10-17 Thread Mark Tinka
On 17/Oct/16 17:03, John Luthcinson wrote: > Hi > > What's the best way to "bleach" the possible QoS (802.1P or DSCP) bits in > CE facing ingress interfaces? The default classifier (ipprec-compatibility) > allows customers to transmit packets to NC queue which is very bad. > > The optimal way wo

[j-nsp] Best way to do QOS bleach

2016-10-17 Thread John Luthcinson
Hi What's the best way to "bleach" the possible QoS (802.1P or DSCP) bits in CE facing ingress interfaces? The default classifier (ipprec-compatibility) allows customers to transmit packets to NC queue which is very bad. The optimal way would be something like this: - classify everything in every