Term
| What is the PHB for Network Control? |
|
Definition
|
|
Term
| What is the PHB for Telephony bearer traffic? |
|
Definition
|
|
Term
| What is the PHB for Telephony Signaling? |
|
Definition
|
|
Term
| What is the PHB for multimedia conferencing? |
|
Definition
|
|
Term
| What is the PHB for Real Time interactive traffic (e.g. gaming, video conferencing) ? |
|
Definition
|
|
Term
| What is the PHB for Multimedia Streaming / video? |
|
Definition
|
|
Term
| What is the PHB for Broadcast video (including live TV)? |
|
Definition
|
|
Term
| What is the PHB for low latency (high -priority) data? |
|
Definition
AF2y
Enhanced best effort. Traffic is assumed to be elastic to adjust to changes in availability. |
|
|
Term
|
Definition
|
|
Term
| What is the PHB for High-Throughput (medium priority) Data? |
|
Definition
|
|
Term
| What is the PHB for Standard Data? |
|
Definition
|
|
Term
| What is the PHB for low-priority data? |
|
Definition
|
|
Term
| What compression algorithm places the most demand on router CPU? |
|
Definition
|
|
Term
| What compression algorithm places the most load on router memory? |
|
Definition
|
|
Term
| Where is QoS info stored on L2 LAN frames? |
|
Definition
For ISL, 3 least significant bits in User Field.
For 802.1Q, User Priority = 3 most significant bits in tag control field |
|
|
Term
| What can an MQC class match on? |
|
Definition
ACL
NBAR
DSCP/Precedence
Other Classes |
|
|
Term
| What does "qos pre-classify" do and where can it be configured? |
|
Definition
Keeps an unencrypted packet in memory so QoS policy can be applied to encrypted traffic at egress.
Command applies to:
• interface tunnel -- GRE & IPIP VPN
• interface virtual-template – L2F & L2TP
• crypto map – IP Sec |
|
|
Term
|
Definition
| QoS Policy Propagation via BGP -- allows you to set ip precedence based on the AS Path, community strings, etc. |
|
|
Term
| Can PBR used to classify and mark? |
|
Definition
| Yes, but only use it if CB-marking is not available. |
|
|
Term
| What does ATM use for QoS? |
|
Definition
| 1 bit Cell Loss Priority (CLP) |
|
|
Term
| How does a switch mark incoming DSCP if auto qos voip cisco-phone is configured and no phone is detected? |
|
Definition
|
|
Term
How does a switch mark incoming DSCP if auto qos voip cisco-phone is configured and CDP indicates a phone is connected? |
|
Definition
QoS markings are trusted. Puts voice, video, routing, STP in priority ingress queue. Puts voice in priority egress queue |
|
|
Term
| What is "auto qos voip trust" for? |
|
Definition
| Connecting to an uplink port. Causes switch to trust CoS on a L2 port, DSCP on a L3 port. |
|
|
Term
| If you enable "auto qos voip" on a Router non-FR serial interface < 768k, what happens? |
|
Definition
| Fragmentation and compression are enabled Encap changes to PPP with PPP Multilink and LFI |
|
|
Term
| What kind of queuing happens in the hardware queue? |
|
Definition
| Hardware is FIFO so SW queues can do work. |
|
|
Term
| If you add up all "bandwidth percent" commands for all classes, what must it be less than. |
|
Definition
| Total of all "bandwidth percent" must not exceed "max-reserved-bandwidth" -- so 75% max by default. |
|
|
Term
| Is LLQ a input or output policy? |
|
Definition
| LLQ service-policies are output only. |
|
|
Term
| How do you configure priority queue bandwidth in a service policy? |
|
Definition
"priority "
"priority percent < % >"
(there is no "priority percent-remaining") |
|
|
Term
| When WRED is configured, what percentage of packets are dropped when the queue depth is between the min and max values? |
|
Definition
Random drops grow with queue depth from 0 to 1/MPD percent.
MPD = Mark Probability Denominator = 10 by default. |
|
|
Term
| With WRED, what is the drop rate when the queue depth exceeds the maximum? |
|
Definition
|
|
Term
| Is WRED precedence based or DSCP based by default? |
|
Definition
"random-detect" == "random-detect precedence"
WRED is precedence based by default. |
|
|
Term
|
Definition
| 12000 version of LLQ. Has 7 round robin and 1 priorirty queue. |
|
|
Term
| What is the following: XpYqZt |
|
Definition
Queue nomenclature.
Xp = number of PQs
Yq = number of regular Qs
Zt = # of WRED threshhold per regular Q |
|
|
Term
| Where is SRR in shared mode queuing used? |
|
Definition
| 3560 ingress or egress queues. SRR is maximally efficient, re-using unused timeslots for waiting traffic. |
|
|
Term
| On an egress queue, what is an internal DSCP value of 1 translated to? |
|
Definition
L2 switches map Internal DSCP to CoS on outbound. Internal DSCPs are also mapped to real DSCP directly (unless there's a mutation map)
Internal DSCP 0 - 7 -> Cos 0
Internal DSCP 8 - 15 -> Cos 1
etc |
|
|
Term
| How is shaped SRR different from shared SSR? |
|
Definition
Shaped SSR does rate-limiting and does not use more than allocated bandwidth even if link is idle.
Shaped is not available on ingress. |
|
|
Term
With the following config, how much weight does the first queue get?
Rack1SW1(config-if)#priority-queue out
Rack1SW1(config-if)#srr-queue bandwidth share 10 20 30 40 |
|
Definition
As much as it needs. When priority queueing is enabled on egress, the PQ is always serviced first and the bandwidth weight for the first queue is ignored. |
|
|
Term
| What is the RSVP message exchange between the gateway and the destination? |
|
Definition
GW -> Dest: PATH
Dest -> GW: RESV
Dest -> GW: PATH
GW -> Dest: RESV
Dest -> Gw: ResvConf |
|
|
Term
| What does "ip rsvp resource-provider none" do? |
|
Definition
| Disables the default WFQ. Configure LLQ as usual after this command (not supported on FR). |
|
|
Term
| How much bandwidth can be reserved by default by RSVP? |
|
Definition
|
|
Term
|
Definition
| Common Open Source Server is an policy decision point for RSVP that switch can consult before making a decision about a reservation. |
|
|
Term
| What is Subnet Bandwidth Manager for? |
|
Definition
| SBM does centralized management of RSVP on LANs. |
|
|
Term
| Can you put a "bandwidth" and a "bandwidth percent" in the same policy map? |
|
Definition
| No, only one bandwidth style is allowed per policy map. |
|
|
Term
What does the following configure?
(config-if)#ip rsvp bandwidth 25000 1000 |
|
Definition
| Allows RSVP to use 1/4 of a 100Mbps link but only allow a single flow to use 1Mbps. |
|
|
Term
| What QoS mechanisms can you use in the inbound direction? |
|
Definition
|
|
Term
| What QoS mechanisms can be used in the outbound direction only? |
|
Definition
shaping
congestion avoidance
queuing |
|
|
Term
| What is the IETF's Default PHB? |
|
Definition
|
|
Term
| What are the building blocks of the DiffServ Traffic Conditioner Block? |
|
Definition
|
|
Term
| What limitations exist when applying queuing to a virtual interface? |
|
Definition
Queuing cannot be applied directly. You must apply a hierarchical policy with shaping and queuing.
policy-map child
class voice priority 512
# Create a "parent" or top-level policy that applies class-based shaping. Apply the child policy as a command under the parent policy since admission control for the child class is done based on the shaping rate for the parent class.
policy-map tunnel
class class-default
shape average 2000000
service-policy child
# Apply the parent policy to the tunnel interface.
interface tunnel0
service-policy tunnel |
|
|
Term
| Can you apply a shaping policy to a tunnel interface? |
|
Definition
Yes.
policy-map tunnel
class class-default
shape average 500000 125000 125000
interface Tunnel0
ip address 130.1.2.1 255.255.255.0
service-policy output tunnel |
|
|
Term
| What is load-interval and how can it be adjusted to account for short bursts? |
|
Definition
Load interval is the length of time for which data is used to compute load statistics.
Default 5 minutes.
To be more reactive to short bursts, shorten the load-interval. |
|
|
Term
| What opcode and port does a TFTP read request use? |
|
Definition
| client send RRQ to TFTP server's port 69 |
|
|
Term
| What does a WRQ to port 69 indicate? |
|
Definition
| A TFTP client is attempting to write a file |
|
|
