Data Center, Professional 試験
Question No : 1
Which two statements describe a VXLAN network identifier (VNI)? (Choose Two)
Question No : 2
You created an IP Fabric topology as shown in the exhibit and are having connectivity problems.
What is causing these problems?
Question No : 3
You must manually configure VXLAN on a QFX5100.
Which protocol is required to perform this task?
Question No : 4
Which statement is correct about EVPN Type 2 routers?
Question No : 5
Which two technologies provide Layer 2 services across data centers in two remote locations? (Choose two.)
Question No : 6
Which protocol replicates forwarding information between MC-LAG peers?
Question No : 7
Which statement is true for an IP Fabric to be called a Clos IP Fabric?
Question No : 8
What are two valid types of VXLAN signaling? (Choose two.)
Question No : 9
You are configuring VXLAN on a network of QFX5100s, and are using the set vlans vlan100 vxlan unreachable-vtep-aging-timer 600 command.
What does this command do?
NOT A C This answer is ambiguous and vague about which MACs are involved.
B C this answer is most detailed and accurate.
C C False; The aging timer refers to the remote VTEP MAC, not the MACs learned *via* that VTEP.
D C False; The aging timer refers to the remote VTEP MAC after MACs learned *via* that VTEP age out; not locally-learned MACs.
Question No : 10
You have an EBGP-based IP Fabric.
Why do you need an export policy on each leaf in this scenario?
A C yes, remote IP prefixes will not be known inside the local autonomous system unless learned via eBGP with remote peer.
NOT B C full mesh not required; would be required in iBGP without RR
C C yes, local IP prefix will not be known inside the peer autonomous system unless advertised.
NOT D C RR only used in iBGP
Question No : 11
You created an IP Fabric based on IBGP. You noticed that ECMP load sharing is not working.
In this scenario, what is the problem?
NOT A C iBGP means by definition the same AS!
B C Because route-reflectors only advertise the best and active route, configuration changes are required to export multiple routes (where applicable); there are a few ways to do this C ‘add path/multipath’ or multiple RRs or multiple sessions with different policies, etc.
NOT C C False; ECMP works on any IP path regardless of interface speed.
D C Route reflector clients must also support ECMP in their import policies and their forwarding policies.
Question No : 12
Which three BGP messages pass EVPN information between devices? (Choose three.)
A C BGP open messages contain information used to negotiate the establishment of sessions but do not contain route information itself. They may contain information such as ‘I can exchange EVPN AFI x/SAFI y NLRI with you’, but not EVPN information itself.
B C BGP update messages can contain EVPN NLRI information. This is the ‘routing information’ itself.
NOT C C BGP notification messages indicate error conditions and contain basic error codes but no routing information. Technically they could refer to an error relating to EVPN, however this is not a good fit for the question being asked.
NOT D C BGP keepalive messages are always 19 bytes and uniform; no NLRI information.
E C A BGP speaker can send a ROUTE-REFRESH message to request the receiver re-sends their previously-sent BGP updates which, as per C, can include EVPN NLRI information.
Question No : 13
You are deploying an IP Fabric in your new data center. You need to ensure that your servers have multiple active/active links for redundancy and load balancing.
What are two methods to accomplish this goal? (Choose two.)
NOT A because EVPN is specifically designed to handle multi-homed access redundancy, and mobility, and load-balancing is possible. I would not choose this though because EVPN is primarily a DCI protocol C for interconnecting datacenters C and is not typically considered an access layer protocol facing the servers. It is possible to use EVPN inside a single datacenter, in which case it would be used between the access switches and the distribution/core switches. In the spirit of the question however, this does not face the servers directly so I exclude it.
B because MSTP allows for multiple spanning trees, which allows for use of all inter-switch links by, for example, aggregating VLANs into spanning trees C eg: VLANs 1-100 in one tree, 101-200 in another. One spanning tree is active on a given link, the other is inactive. In this way load-balancing is achieved. In case of link failure, all traffic will fall to the remaining link(s).
C because MC-LAG allows for link aggregation across multiple chasses -> permits load-balancing and provides redundancy.
NOT D because PIM-SM is a control-plane signaling protocol for multicast which enables routers to join and prune their membership to given multicast distribution trees. It has little directly to do with the redundancy and load-balancing of (multicast) traffic receivers.
Question No : 14
You are establishing an IP Fabric where your QFX10002-72Q spine devices are connecting to leaf devices. Each leaf is a Virtual Chassis comprised of ten EX4300s.
What is the maximum number of leaf devices supported in this scenario?
Question No : 15
Referring to the exhibit, what is the MAC address of the IRB interface on router PE1?