3850 QoS question

Hello,
I am trying to upgrade from 3750 to 3850. Following is my outbound WTD qos config on 3750:
mls qos srr-queue output cos-map queue 1 threshold 3 5
mls qos srr-queue output cos-map queue 2 threshold 1 2 4
mls qos srr-queue output cos-map queue 2 threshold 2 3
mls qos srr-queue output cos-map queue 2 threshold 3 6 7
mls qos srr-queue output cos-map queue 3 threshold 3 0
mls qos srr-queue output cos-map queue 4 threshold 3 1
mls qos srr-queue output dscp-map queue 1 threshold 3 46
mls qos srr-queue output dscp-map queue 2 threshold 1 16 18 20 22 32 34 36 38
mls qos srr-queue output dscp-map queue 2 threshold 2 24 26 28 30
mls qos srr-queue output dscp-map queue 2 threshold 3 48 56
mls qos srr-queue output dscp-map queue 3 threshold 3 0
mls qos srr-queue output dscp-map queue 4 threshold 1 8
mls qos srr-queue output dscp-map queue 4 threshold 3 4 10 12 14
mls qos queue-set output 1 threshold 2 70 80 100 100
mls qos queue-set output 1 threshold 4 40 100 100 100
What will be the equivalent config for the 3850? I am not able to find the right documentation for 3850.
Following is what I created but I believe it wouldnt handle WTD.
class-map match-any EF
match dscp ef
class-map match-any CS3AF41
match dscp cs3 af41
class-map match-any AF21
match dscp af21
class-map match-any AF11
match dscp af11
class-map match-any Default
match dscp default
policy-map QOS-OUTBOUND
class EF
   priority level 1
class CS3AF41
   priority level 2
class AF21
   bandwidth remaining percent 20
   shape average percent 100
class AF11
   bandwidth remaining percent 20
   shape average percent 100
class Default
   bandwidth remaining percent 60
   shape average percent 100
Also, when I try to apply this to an interface, it accepts it but its not visible under show runn. Am I missing anything?
I would appreciate any help with this.
Thank you.

Hi,
3850 is not supporting MLS QoS (so any QoS commands start with mls keyword is not applicable in this switch platform- even thoug they are accepted).
It is a MQC QoS based switch platform (like 6500,4500). There are few Queing models (1P7Q3T, 2P6Q3T for wired &  2P2Q for wireless) supported in this platform. Please refer below post for more detail
http://mrncciew.com/2013/12/23/3850-qos-part-2-queuing-models/
Based on this you need to derive your config changes. As you can see in the above "queue limit x " is the command used to configure WTD for non-priority queues (ie Q2 to Q6)
Regarding running config, you can use "show running-config all" command to see all the configuration lines of your switch including default configs
HTH
Rasika
***Pls rate all useful responses ****

Similar Messages

  • 3850 QoS class-map match-all?

    I would like to create a QoS marking policy that re-marks packet to CS5 if the inbound traffic is SIP *and* if it is marked CS3 when it comes in.  I would have expected the configuration listed below will work.  I only found out when I tried to apply the config that, unlike other IOS devices, "class-map match-all" does not exist in 3850 3.3.x code.  It can only do "class-map match-any" Can anyone suggest a work-around config for 3850 to achieve the same end result?
    ip   access-list extended ACL-QOS-SIP
    permit tcp any range 5060 5061 any
    permit tcp any any range 5060 5061
    ip   access-list extended ACL-QOS-CS3
    permit ip any any dscp cs3
    class-map   match-all CM-QOS-CS5
      match access-group name ACL-QOS-CS3
      match access-group name ACL-QOS-SIP
    policy-map   PM-QOS-MARKING
    class CM-QOS-CS5
         set ip dscp cs5
    Any suggestions would be appreciated.

    jlkeys,  below is configuration I ended up using to resolve the issue:
    ip access-list extended ACL-QOS-SIP
     permit tcp any range 5060 5061 any dscp cs3
    class-map match-any CM-QOS-CS5
      match access-group name ACL-QOS-SIP
    policy-map   PM-QOS-MARKING
     class CM-QOS-CS5
       set ip dscp cs5

  • Basic QoS questions

    Hello,
    Out company is finally going to install VOIP, they will be using something called Natterbox, but I've been asked to add QoS to our Cisco switches and routers in various locations I will also need to add a second VLAN for voice I believe which is ok.
    How can we configure QoS so VOIP is treated first then video streaming then video?
    As we will need to add a second VLAN, so I assum all ports to phone swill be truk ports?
    Thanks

    Andy,
    To start, you need to understand how the soft phone and desk phones mark voice traffic. Cisco switch and Routers perform QoS by trusting, understanding or marking traffic using DSCP, PHB and port numbers.
    Hence questions you need to ask are
    1. Do the softphones mark their voice and video traffic? If they do how do they do it?
    2. What ports are used for voice traffic?
    3. what ports are used for signalling traffic
    4.What ports are used for video traffic if its differs from ports used for voice.
    Once these facts are understood, you can beging to look at how to configure your switches and routers based on this parameters
    Please rate all useful posts
    "opportunity is a haughty goddess who waste no time with those who are unprepared"

  • QoS Questions for 3750 Switches

    2x3750 switches are stacked and we are trying to simulate traffic congestion at the UTP ports by using Smartbit 6000C. The objective of the test is to see if the QoS setting works in reality even though we see from Wireshark that the packets are marked with DSCP for voice traffic.
    Setup is as follows :
    Smartbit<->Avaya IP Phone<->3750 switches<->6509 switch.
    Please note that the configuration is set on the 3750 switch port as well as trusted on the Cat 6 switch port. However, when I set to continuous traffic with byte size of 64. Even though its a 100Mbps port, the Avaya IP Phone is already acting weird with hanged symptom. Just side note is that performing "show mls qos inter gi2/0/7 statistic" shows that data and voice traffic are marked on the different priority which seemed correct.
    1) Is this the right way to test? If not, what should be the correct way?
    2) The port that's connected to Smartbit is configured and it seemed that with the continuous traffic, even other IP Phones are hanged even though I have set Smartbit to hit on the IP address of the CAT6 Switch port. This is not normal right as this is supposed to be unicast traffic. Any idea what could be the reason?

    Hello Brandon,
    I understand your concern and how you want to test, but with the VoIP services you need to understand that there are 2 points (telephones if you want) involved. Your local one, where you might have taken all the necessary steps to protect and prioritize your voice traffic, and the oposite end which also need to have the voice packets prioritized.
    Now, from your description, I understand that the packets (voice and data) marked correctly (I believe on C3750), but that's not enough. You need to use CBWFQ together with LLQ to give priority to the Voice traffic over data in case of congestion. Do you have such configuration? Can you show us some excerpt from it?
    Next, during the testing, you said that your phone hang-up...where you in a call?
    To respond to your questions:
    1. The start is ok, but we need more details. You are pushing traffic from Smarbit, this is your local end, but where is the traffic pushed to (remote end), who is receiving the traffic?
    2. In theory, you shouldn't have any impact over voice if links are 100Mbit, only if you have such a power packet generator that could fill 100Mbit. What do you mean by "This is not normal right as this is supposed to be unicast traffic"? VoIP is also unicast traffic...
    I can see that you are determined to solve this issue, and this is OK as it will help you back with gained knowledge, but I have to warn you that troubleshooting QoS / Voice related problems may be more tricky than you think, as it will involve a strong know-how in these areas.
    We will help, but you have to come back with more precise details.
    Good luck!
    Calin

  • ESW 520 QoS questions

    Hello,
    just to put this out there, I am a noob to the world of Cisco, just got my CCNA, and am now working as an intern.
    Can you set esw 520 switches to use auto QoS?  We are putting the 520's into an environment of more configurable catalyst switches, and they all use auto QoS.  I am basically wondering how to get the QoS on the 520's to play nice with the auto qos on the nicer catalyst switches...
    I have tried just using basic QoS, but that wont work, because when I use the port wizard, I use the macro "ip phone + desktop"
    It insists on setting qos, but fails because voice-map does not exists.  The only way I have found to get the macro to work, is to set QoS to advanced, and create a policy map called "voice-map"  of course, after that the macro works, but then I am left with advanced QoS enabled, and hardly configured...how am I supposed to know how to configure it if the other switches ( catalyst series) are set to auto?
    Im sorry for the seemingly dumb question, but I know next to nothing about QoS other than it is used to prioitize traffic based on data types.

    Hey cole, typically with theses switches the smartports will do the QOS with the port role, however you can manually create the QOS for the switch, by creating a ACL to identify the specific traffic comming through the switch port.  Then Create a class map/ policy map/ and then bind it to the ports your wanting to setup QOS on.

  • RV016 - QoS questions

    Good evening everyone
    I'm a network administrator and I have been struggling with QoS and I will explain why.
    My RV016 is directly connected to the modem and to a gigabit switch as well. there's only one cat6 carrying the traffic to the switch, which is connected to every single device that we have in our network, including IP telephones.
    The RV016 provides two types of QoS which is the prioritazation of an application, where I have already configured to prioritize RTP traffic for voice(ports5000-9000). However, if I prioritize the port which carries all my network traffic with a "High"priority, I BELIEVE it will prioritize and tag every type of traffic that is being passed through the wire, making my application QoS configuration completely useless. I had never taken any QoS courses, so I'm just assuming that this is what is going to happen. .
    My question here is, if I leave out the port that carries all the traffic with a "normal" priority tag instead of "High" and prioritize RTP traffice through application QoS, will that be enough to prioritize RTP traffic? Or do I need to configure each IP Phone to a port and prioritize the port  to high as well to ensure that RTP traffice will be prioritized?
    Any help would be highly appreciated. Thanks in advance!

    Thiago,
    Whenever QOS  is implemented it has certain criteria in which it has to match or it will not apply the QOS  to the frame.
    There is port based, where everything that passes through that port will receive the higher quality of service than the other ports.
    Then there is protocol or ip based QOS, where if the protocol matches that protocol or port number then it receives the QOS.  If it is ip based, it has to match the ip address range you specify to receive the QOS.
    So if you have QOS implemented to match those protocol ports, then only if the traffic matches that criteria verbatim will it be applied.
    In the Switching world, you have to create acl's to match certain criteria to apply qos.  You can say any packet in a certain ip address scheme, matching this protocol port will receive this QOS.
    Hope this helps expain QOS  a little bit for you.

  • UWN QOS questions

    Hello all.
    I've got two questions related to wireless QOS in a UWN world:
    1) A QOS profile is associated with a WLAN. Is it possible to diferentiate traffic within a single WLAN? Can I assign the gold profile to one DSCP value, and the silver profile to another DSCP - all within the same WLAN? If not, how do others differentiate between applications on a mobile device (email vs. softphone)?
    2) Is it possible for the LWAPP AP to remap or set incoming DSCP - the classification and marking functions of QOS (aka establish the QOS classification boundry on the radio). If not, I'm concerned about the WLC maintaining potentially invalid DSCP in the LWAPP header on the way to the controller. How do others ensure they aren't oversubscribing a priority queue on a WAN link from an H-REAP AP to a central site controller?

    Unfortunately, I'm almost positive the answer to both questions is NO.
    From the document you referenced -
    "The controller does not apply its own QoS. The QoS support on the WLC gives the WLC the ability to apply the same priority that is set on the wire (or application). Therefore, the only action a WLC or AP will do is copy the value of the original packet to the outer header of the LWAPP packet."
    For question 1, to use different QOS profiles, we must associate clients to different WLANs. The result is that we are limited to device-specific QOS, as opposed to application-specific QOS. If we want VoIP or Video over wireless, the endpoints (e.g. VoIP handsets) must connect to their own SSID. Softphone traffic can't be preferred over other PC traffic.
    For question 2, this means we need to be concerned about the LWAPP AP maintaining potentially invalid DSCP in the LWAPP header on the way to the controller (and vice-versa). The problem comes in that this may oversubscribe a priority queue on a WAN link from a remote AP to a central site controller, unless you are policing at L3 ingress (which isn't without problems too). The way around this is to conditionally clear DSCP on ingress (e.g. not trusting the controller or AP uplink).

  • 5760 MC with 3850 MA question

    With access points connected to a 3850 configured as a mobility agent of a 5760 mobility controller, on which device are the WLAN's and AP groups defined? I have a test environment set up in preparation for changing the role of several 3850 MC's to MA's after we install a 5760. I changed the role of a previously configured 3850 from MC to MA, and then defined the 5760 as the MC on it after it rebooted. I then added the 3850 as a member of a peer group on the 5760. The data and control path between the two came up and everything looked good. Then I connected a 3702 to the 3850. It shows as registered on the 3850 with the show ap sum command, and on the 5760 it appears as learnt from mobility agent with the show wireless mobility ap list command. The access point however it is in the ap group defined on the 3850, and advertising the wlans defined on the 3850. I tried to change the 3850 connected access points ap group from the 5760 to one of the ap groups defined on the 5760, but the command fails with the error invalid name or ap. It seems like ap commands cannot be run against access points that are not directly connected. So where are the interfaces, Wlans and access point groups defined in the converged access architecture? On the 3850 MA, or the 5760 MC? Or am I having this problem because I only changed the role of a previously configured 3850 MC without removing the wlans, ap groups, and client interfaces? I have read several convereged access docs, but am having trouble understanding how to configure each component. Please ask questions if you need more information to respond.

    You will want the same AP groups to exist on all of your 3850s and 5760s.  The AP only physically resides on the directly connected switch and currently you must configure each piece separately.  Each 3850 stack is essentially still a WLC (configured as MA), but the 5760 now handles client roaming, RRM, WIPS, ect.  So while you can configure the AP group on the 5760 the access points will continue to stay within the group specified on the 3850 it is directly connected to.
    You will still have to configure everything on the 3850s as before.  The difference now is that you can now scale to a larger network.  
    It is a pain now, but in the future you will configure almost everything on the 5760 and it will push the changes out to the 3850/3650 switches for you.
    I hope this helps.
    John

  • 3850 Stack Question. I'm new to stacks

    So, I have two 3850's I want to put in a closet.  Do I stack them or just leave them stand alones...  My plan was to connect each to the core switch using fiber and users on each switch.
    When I conenct the switches using the stackwise cable, I have an active and standby...  Is there a way to make the standby a member?  is the standby in true "standby" mode?  Can I use the ports on the standby switch?
    SO many questions   Apppriecte any help..
    Thanks

    Disclaimer
    The  Author of this posting offers the information contained within this  posting without consideration and with the reader's understanding that  there's no implied or expressed suitability or fitness for any purpose.  Information provided is for informational purposes only and should not  be construed as rendering professional advice of any kind. Usage of this  posting's information is solely at reader's own risk.
    Liability Disclaimer
    In  no event shall Author be liable for any damages whatsoever (including,  without limitation, damages for loss of use, data or profit) arising out  of the use or inability to use the posting's information even if Author  has been advised of the possibility of such damage.
    Posting
    BTW, when you stack your pair, ideally, you still want an uplink from each switch.  Depending on L2/L3 configuration, on both your stack and core device, you should be able to use the bandwidth from both links to/from the stack too.  Also, if possible, try not to have the two links terminate on same line card on your core device, to avoid that also being a single point of failure.

  • SG300-28P QoS Question

    Hello,  I have SG300-28Ps as the PSE's for my IP telephone system.  The phones are tagging their voice packets as DSCP 46 as directed by auto voice vlan. The QoS settings on the switch are at default - Basic Mode, Trust DSCP, strict priority, etc.
    On the PBX itself, DB programming allows me to program the 'Type of Service' for the voice packets. The recommended value in the manual was 184 which makes sense, as this decimal value for ToS corresponds to DSCP 46, CoS 5, etc.
    The question comes though, do I need to change the trust mode on the switch? I'm not real clear on the differences between them.
    Regards,
    -Brayton

    Hi Brayton,the trust mode doesn't need to be changed. 802.1p specifies a 3 bit field called a  PCP within the etherner frame header when using tagged vlan frames. This will contain a class of service priority.
    The CoS is able to map to DSCP values. The DSCP has a 6 bit field called diffserv (differentiated service). CoS values are able to be mapped to DSCP values. Video is generally CoS 4 while voice is generally CoS 5. Within the SX300 you are able to manually set the mapping to any value you'd like. With trust mode, the switch will basically accept and agree with whatever the tagged ethernet frame contains. Without trust mode, the switch will remark the packet based on the PCP and DiffServ value to fit in to the different categories.
    -Tom
    Please mark answered for helpful posts

  • Catalyst 3850 QoS police

    Hello,
    Here is the config for Catalyst 3560 found under the link below.
    I would like to do same setting on Catalyst 3850.
    http://itknowledgeexchange.techtarget.com/network-engineering-journey/how-to-configure-per-vlan-qos-in-cisco-3550-and-3560/
    mls qos
    interface fa0/2
    mls qos vlan-based
    class-map INT
    match input-interface fa0/2
    policy-map NESTED_POLICE
    class INT
    policy 12800 1600 exceed-action drop
    class-map HTTP
    match protocol http
    policy-map PARENT_MARK
    class HTTP
    set dscp af11
    service-policy NESTED_POLICE
    interface vlan 10
    service-policy input PARENT_MARK
    But commands like "mls qos", "mls qos vlan-based" and "match input-interface " doesn't work on 3850.
    There is no helpful Cisco manual for it.
    Could anyone help me?
    Thanks in advance,
    Taro

    Hello Paul,
    Thank you for the attention.
    Here is the information.
    #sh ver
    Cisco IOS Software, IOS-XE Software, Catalyst L3 Switch Software (CAT3K_CAA-UNIVERSALK9-M), Version 03.02.01.SE RELEASE SOFTWARE (fc1)
    Technical Support: http://www.cisco.com/techsupport
    Copyright (c) 1986-2013 by Cisco Systems, Inc.
    Compiled Wed 20-Mar-13 17:10 by prod_rel_team
    Cisco IOS-XE software, Copyright (c) 2005-2013 by cisco Systems, Inc.
    All rights reserved.  Certain components of Cisco IOS-XE software are
    licensed under the GNU General Public License ("GPL") Version 2.0.  The
    software code licensed under GPL Version 2.0 is free software that comes
    with ABSOLUTELY NO WARRANTY.  You can redistribute and/or modify such
    GPL code under the terms of GPL Version 2.0.
    (http://www.gnu.org/licenses/gpl-2.0.html) For more details, see the
    documentation or "License Notice" file accompanying the IOS-XE software,
    or the applicable URL provided on the flyer accompanying the IOS-XE
    software.
    ROM: IOS-XE ROMMON
    BOOTLDR: C3850 Boot Loader (C3850-HBOOT-M) Version 1.1, RELEASE SOFTWARE (P)
    SW01 uptime is 21 weeks, 6 days, 14 hours, 27 minutes
    Uptime for this control processor is 21 weeks, 6 days, 14 hours, 30 minutes
    System returned to ROM by reload at 22:27:58 JST Wed Jan 8 2014
    System restarted at 22:27:52 JST Wed Jan 8 2014
    System image file is "flash:packages.conf"
    Last reload reason: Reload command
    This product contains cryptographic features and is subject to United
    States and local country laws governing import, export, transfer and
    use. Delivery of Cisco cryptographic products does not imply
    third-party authority to import, export, distribute or use encryption.
    Importers, exporters, distributors and users are responsible for
    compliance with U.S. and local country laws. By using this product you
    agree to comply with applicable laws and regulations. If you are unable
    to comply with U.S. and local laws, return this product immediately.
    A summary of U.S. laws governing Cisco cryptographic products may be found at:
    http://www.cisco.com/wwl/export/crypto/tool/stqrg.html
    If you require further assistance please contact us by sending email to
    [email protected].
    License Level: Ipservices
    License Type: Permanent
    Next reload license Level: Ipservices
    cisco WS-C3850-24T (MIPS) processor with 4194304K bytes of physical memory.
    Processor board ID FOC1717V01B
    24 Virtual Ethernet interfaces
    56 Gigabit Ethernet interfaces
    8 Ten Gigabit Ethernet interfaces
    2048K bytes of non-volatile configuration memory.
    4194304K bytes of physical memory.
    250456K bytes of Crash Files at crashinfo:.
    250456K bytes of Crash Files at crashinfo-2:.
    1609272K bytes of Flash at flash:.
    1609272K bytes of Flash at flash-2:.
    0K bytes of Dummy USB Flash at usbflash0:.
    0K bytes of Dummy USB Flash at usbflash0-2:.
    0K bytes of  at webui:.
    Base Ethernet MAC Address          : 44:ad:d9:6d:4e:00
    Motherboard Assembly Number        : 73-12238-06
    Motherboard Serial Number          : FOC17163HB8
    Model Revision Number              : B0
    Motherboard Revision Number        : D0
    Model Number                       : WS-C3850-24T
    System Serial Number               : FOC1717V01B
    Switch Ports Model              SW Version        SW Image              Mode
         1 32    WS-C3850-24T       03.02.01.SE       cat3k_caa-universalk9 INSTALL
         2 32    WS-C3850-24T       03.02.01.SE       cat3k_caa-universalk9 INSTALL
    Switch 02
    Switch uptime                      : 21 weeks, 6 days, 14 hours, 31 minutes
    Base Ethernet MAC Address          : 20:bb:c0:01:86:80
    Motherboard Assembly Number        : 73-12238-06
    Motherboard Serial Number          : FOC17163HCM
    Model Revision Number              : B0
    Motherboard Revision Number        : D0
    Model Number                       : WS-C3850-24T
    System Serial Number               : FOC1717V01K
    Configuration register is 0x102
    SW01#sh sdm prefer
    Showing SDM Template Info
    This is the Advanced template.
      Number of VLANs:                                 4094
      Unicast MAC addresses:                           32768
      Overflow Unicast MAC addresses:                  512
      IGMP and Multicast groups:                       8192
      Overflow IGMP and Multicast groups:              512
      Directly connected routes:                       32768
      Indirect routes:                                 8192
      Security Access Control Entries:                 3072
      QoS Access Control Entries:                      2816
      Policy Based Routing ACEs:                       1024
      Netflow ACEs:                                    1024
      Input Microflow policer ACEs:                    256
      Output Microflow policer ACEs:                   256
      Flow SPAN ACEs:                                  256
      Tunnels:                                         256
      Control Plane Entries:                           512
      Input Netflow flows:                             8192
      Output Netflow flows:                            16384
    These numbers are typical for L2 and IPv4 features.
    Some features such as IPv6, use up double the entry size;
    so only half as many entries can be created.

  • Nexus and 3850 QoS

    Dear all,
    I am having a problem with implementing QoS for a 3rd Party application. The vendor requests that some of this traffic needs special markings with specific cos/dscp values. Also we need to have QoS for our Cisco IP Phones.
    The devices ( phones and the 3rd Party devices) are hooked up at the 3850 (running IP Base) switches which are connected via port channel to 2 Nexus 5596UP which are connected via vPC.
    If i execute the 
    show policy-map int t1/1/4 or show policy-map int e1/1 on the switches, I can see that there are match packets ( matched by any) but 0 by the UDP ports.
    Also the the 3rdParty Clocks are not synchron which is an indicator that the QoS is not working. If the two clocks are hooked up at the same 3850 switch they sync. 
    I am sure I miss something, but I really do not know what. The goal is to have QoS for the Cisco IP phones and also the 3rd Party devices.
     Any help would be much appreciated.
    Here the 3rd Party requests and configurations I did on the switches:
    Protocol: PTP
    Destination: 224.0.1.129
    Dest.Port: UDP319,320
    DSCP: CS7
    CoS: 56
    Protocol: Unicast Streaming
    Dest.Port: UDP 14336-14600
    DSCP: EF
    CoS: 46
    Protocol: Broadcast Streaming
    Dest. Port UDP 4321
    DSCP: EF
    CoS: 46
    I created the following Class Maps, ACLs and Policy Maps on the Nexus Switches for 3rd Party and Voice:
    ip access-list ACL-Dante-Audio
      10 permit udp any any eq 4321
      20 permit udp any any range 14336 14600
      30 permit ip any any
    ip access-list ACL-Dante-PTP
      10 permit udp any any eq 319
      20 permit udp any any eq 320
      30 permit ip any any
    class-map type qos match-any VOICE
      match dscp 46
      match cos 5
    class-map type qos match-any Dante-PTP
      match access-group name ACL-Dante-PTP
    class-map type qos match-any Dante-Audio-PTP
      match access-group name ACL-Dante-Audio
    policy-map type qos Global-QoS-IN
      class Dante-PTP
        set dscp 56
      class Dante-Audio-PTP
        set dscp 46
      class VOICE
        set qos-group 5
    system qos
      service-policy type qos input Global-QoS-IN
    The following settings are configured on the 3850's:
    ip access-list extended ACL-Dante-Audio
     permit udp any any eq 4321
     permit udp any any range 14336 14600
     permit ip any any
    ip access-list extended ACL-Dante-PTP
     permit udp any any eq 319
     permit udp any any eq 320
     permit ip any any
    class-map match-any Dante-PTP
    match access-group name ACL-Dante-PTP
    class-map match-any Dante-Audio-PTP
     match access-group name ACL-Dante-Audio
    class-map match-any VOICE
     match dscp ef 
     match cos  5 
    policy-map Output-Qos-Policy
     class Dante-PTP
      set dscp cs7
     class Dante-Audio-PTP
      set dscp ef
     class VOICE
    interface TenGigabitEthernet1/1/4
     description Uplink_Core1_e1/1
     switchport mode trunk
     switchport nonegotiate
     auto qos trust dscp
     channel-group 1 mode passive
     service-policy input AutoQos-4.0-Trust-Dscp-Input-Policy
     service-policy output Output-Qos-Policy
     ip dhcp snooping trust
    interface TenGigabitEthernet2/1/4
     description Uplink_Core2_e1/1
     switchport mode trunk
     switchport nonegotiate
     auto qos trust dscp
     channel-group 1 mode passive
     service-policy input AutoQos-4.0-Trust-Dscp-Input-Policy
     service-policy output Output-Qos-Policy
     ip dhcp snooping trust
    Thanks in advance
    Thorsten

    After a few tries we found solution.
    Solution is customizing cisco av pair, it must looks like:
    cisco-av-pair=shell:roles*"network-admin admin"

  • Newb VoIP End to End Qos Questions

    I'm at a site that has Layer 3 Cisco switches and routers end to end. Cisco IP phones plug directly into Layer 3 switches with various routers in the core.
    1) If all connections are Layer 3 do I need to concern myself with trusing COS? I would trust DSCP in this case yes?
    2) Cisco phones by default send voice rtp audio with DSCP marked as EF, is that correct? If not, I know how to create policy-maps and reclassify it. My thought was it is already classified as high as possible.
    3) I've read that QoS must be end to end or you don't really have a full implementation, as in the one device your traffic passes through that is not performing QoS will likely be running FIFO or WFQ, which do not give precedence to class based traffic. So in theory I need to implement CBWFQ or LLQ on every Layer 3 device end to end that RTP audio will touch. Do I have this correct?
    4) If the voice traffic has the potential to travel to most Layer 3 devices in the network as it travels between the Call Manager and phones, I need to apply QoS outbound on each Layer 3 interface that is likely to be in a voice path yes?
    Sorry, I'm a newb, I understand it in theory but I still hit the wall on occassion with my light saber while blindfolded.
    Thanks, Mike

    Hi
    1) Yep - no trunk (dot1q/ISL) = no CoS as it's in the trunk header, so use DSCP.
    2) Yep, traffic will be marked. You can trust it, or trust it conditionally in a service policy and police it.
    3) Yep, more or less. It's very important that on any switches that have QoS enabled, you have proper trust states set on all the ports, otherwise the DSCP will be remarked to 0. This is worse that not particularly prioritizing voice on a particular device, as no other devices further down the chain will see any DSCP markings to act on.
    4) Erm.. kind of. If it's a layer 3 link on a switch, then you don't configure it as you would on a router. The QoS is done in hardware, so your common/garden autoqos type config should see you right if you really don't know what you are doing. If you use auto qos voip trust, make sure you set it to mls qos trust dscp after as it will probably default to CoS. Best to keep it consistent...
    Regards
    Aaron
    Please rate helpful posts...

  • QoS questions SG200

    If I don't have any VOIP devices, should I just leave all the QoS settings to default? We have an AirPlay speaker on our network, so could there be benefits in raising the QoS priority so that the audio streams smoothly to this device?
    Cheers,
    Max

    Hi MaxHodges, the default QoS will apply to only tag VLAN packet from 802.1p. Otherwise the traffic is treated first in first out basis and switching at hardware speed.
    Unless you have a significant amount of traffic that causes congestion there shouldn't be any problem. Separating traffic on a VLAN may logically separate the packets but it wouldn't separate the physical wire usage. So if everything is working as intended then there shouldn't be any need to do anything.
    If you are experiencing problems with things like choppiness, you may want to troubleshoot from a congestion point of view such as removing all other traffic off the link and see if its clear or continues to have a problem. At that juncture, if you determine that with additional traffic load causes a problem then it may be prudent to think about a VLAN or reducing the amount of traffic on the un-aggregated link or possibly adding another uplink in to your network.
    If the choppiness/quality doesn't clear with just your music box going then chances are it's not the switch or if it's the switch then it'd likely be a wiring issue or doubtfully the switch itself being flaky.

  • Qos questions

    Hi all,
    need some advice here ..
    1) for QoS, which one better, configure inside CE or PE
    2) Can we do QoS based on time/date. ie end of the month or after 6 pm, e-mail become silver ..other than that, e-mail traffic is precendence 0. And also bandwidth allocation. That means, middle of the month, gold is 40% and end of the month, gold is 15 %.
    If yes, any good link/url for this. I think it can be done but just wanna get opinion/advice from u all.
    rgds,
    --mdshah

    You can use the modular QoS CLI to achieve this. You can create time based ACLs and use them in the policy-maps to set IP precedence or other parameters like MPLS EXP bits or DSCP values, based on which QoS treatment can be given to the traffic. Check out the 'Classification' and 'Modular QoS CLI' setions in this link :
    http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122cgcr/fqos_c/index.htm

Maybe you are looking for