Skip to content

Professional Cloud Network Engineer · Domain 4: Configuring and implementing hybrid and multicloud network interconnectivity · 16% of the exam

Task 4.3: Configuring Cloud Router.

BGP settings such as ASN, route priority and MED, link-local addresses and authentication, BFD, custom advertised and custom learned routes, and legacy or standard best path selection.

Study it

  • Cloud Router: ASN, route priority and MED, BFD, custom advertised and custom learned routes, best path selection

    Lesson coming

Sample questions

Try each one before opening the answer. Every option is explained, with the Google Cloud documentation page that proves it.

Question 1 · choose 1

A VPC network uses global dynamic routing. Cloud Router cr-central in us-central1 has BGP sessions on attachment A, and cr-east in us-east4 has sessions on attachment B; all sessions use the default base priority. The data center's routers choose the path with the lowest MED. The team wants the data center to send traffic for a us-central1 subnet over attachment B, using attachment A only if B fails. Which base priorities achieve this whatever the region-to-region cost?

  1. A0 on the cr-east sessions, 200 on the cr-central sessions
  2. B50 on the cr-east sessions, 9000 on the cr-central sessions
  3. C50 on the cr-east sessions, 10050 on the cr-central sessions
  4. DA custom learned route priority of 0 on the cr-east sessions only
Show the answer and why
  • A0 on the cr-east sessions, 200 on the cr-central sessions

    Incorrect

    cr-east advertises the us-central1 subnet with 0 plus a region-to-region cost of at least 201, which is always higher than 200.

  • B50 on the cr-east sessions, 9000 on the cr-central sessions

    Incorrect

    Region-to-region costs can reach 9999, so cr-east's MED for the subnet can be as high as 10049, above 9000.

  • C50 on the cr-east sessions, 10050 on the cr-central sessions

    Correct

    cr-east's MED for the subnet is at most 50 + 9999 = 10049, below cr-central's 10050, so B is preferred and A remains as backup.

  • DA custom learned route priority of 0 on the cr-east sessions only

    Incorrect

    Custom learned route priority affects routes Google Cloud uses toward the data center, not the MED the data center sees.

The MED Cloud Router advertises for a subnet equals the session's base priority, plus a region-to-region cost when the subnet is in another region under global dynamic routing; cr-central advertises its own region's subnet with the base priority alone. The cost is between 201 and 9999 and can't be changed, so cr-east's MED for the subnet lies between its base priority plus 201 and its base priority plus 9999, and only a cr-central priority above that upper bound keeps B preferred for every cost.

Question 2 · choose 1

Behind the data center's BGP router sits a legacy lab network, 192.168.50.0/24, that the router can't advertise over BGP. VMs must reach it through the existing HA VPN tunnels, and the route should disappear from the VPC network whenever the BGP session on its tunnel goes down. What should the network engineer configure?

  1. AA static route for 192.168.50.0/24 whose next hop is one of the HA VPN tunnels
  2. BA custom learned route for 192.168.50.0/24 on the Cloud Router BGP sessions of the tunnels
  3. CA custom advertised route for 192.168.50.0/24 on the Cloud Router BGP sessions of the tunnels
  4. DA policy-based route for 192.168.50.0/24 whose next hop is the HA VPN gateway
Show the answer and why
  • AA static route for 192.168.50.0/24 whose next hop is one of the HA VPN tunnels

    Incorrect

    A static route's VPN tunnel next hop must be a Classic VPN tunnel, not an HA VPN tunnel.

  • BA custom learned route for 192.168.50.0/24 on the Cloud Router BGP sessions of the tunnels

    Correct

    Custom learned routes are manually defined routes applied to a BGP session, and they are withdrawn when that session goes down.

  • CA custom advertised route for 192.168.50.0/24 on the Cloud Router BGP sessions of the tunnels

    Incorrect

    Custom advertisements tell the data center about ranges in Google Cloud; they create no route in the VPC network.

  • DA policy-based route for 192.168.50.0/24 whose next hop is the HA VPN gateway

    Incorrect

    A policy-based route sends packets to an internal passthrough Network Load Balancer, not to a VPN gateway.

Cloud Router normally learns egress routes from its BGP peers. Custom learned routes let you add such routes by hand to a BGP session; they are the only way to define non-BGP routes for VLAN attachments and HA VPN tunnels, and they follow the session's state.

Question 3 · choose 1

One Cloud Router manages BGP sessions over two Dedicated Interconnect VLAN attachments. The data center team asks for a different Google-side ASN on each session so that it can tell the attachments apart in its BGP tables. What should the network engineer answer?

  1. AEach BGP session can be given its own local ASN on the same Cloud Router
  2. BA Cloud Router uses the same ASN for all its BGP sessions, so a different ASN needs another Cloud Router
  3. CThe Google-side ASN is fixed by Google and can't be chosen at all
  4. DEach VLAN attachment carries its own ASN, independent of the Cloud Router
Show the answer and why
  • AEach BGP session can be given its own local ASN on the same Cloud Router

    Incorrect

    A Cloud Router uses one ASN for all of its BGP sessions.

  • BA Cloud Router uses the same ASN for all its BGP sessions, so a different ASN needs another Cloud Router

    Correct

    Each Cloud Router uses the same ASN for all of its sessions; separate ASNs mean separate Cloud Routers.

  • CThe Google-side ASN is fixed by Google and can't be chosen at all

    Incorrect

    You choose the Cloud Router's ASN when you create it.

  • DEach VLAN attachment carries its own ASN, independent of the Cloud Router

    Incorrect

    The Google-side ASN of a session comes from its Cloud Router, not from the attachment.

You set a Cloud Router's ASN when you create it, and that Cloud Router uses the same ASN for every BGP session it manages, whether over VLAN attachments or VPN tunnels.

Question 4 · choose 1

BFD is enabled on a Cloud Router BGP session with minimum transmit and receive intervals of 300 ms and a multiplier of 5. The peer router keeps its intervals at 1,000 ms with a multiplier of 5. Cloud Router still takes 5 seconds to declare the session down after the link fails. Why?

  1. AThe peer's 1,000 ms intervals govern the negotiated BFD timing
  2. BCloud Router ignores custom BFD intervals and always uses 1,000 ms
  3. CBFD can only detect failures on HA VPN tunnels, not on VLAN attachments
  4. DBFD detection always adds the BGP hold timer to the BFD interval
Show the answer and why
  • AThe peer's 1,000 ms intervals govern the negotiated BFD timing

    Correct

    Cloud Router's detection time is the peer's agreed transmit interval times the peer's multiplier: 1,000 ms x 5.

  • BCloud Router ignores custom BFD intervals and always uses 1,000 ms

    Incorrect

    The intervals are configurable; the negotiation with the peer decides the result.

  • CBFD can only detect failures on HA VPN tunnels, not on VLAN attachments

    Incorrect

    BFD runs inside BGP sessions over VLAN attachments too.

  • DBFD detection always adds the BGP hold timer to the BFD interval

    Incorrect

    BFD replaces BGP-based detection, which takes 60 seconds by default.

BFD negotiates each direction's agreed transmit interval as the higher of the sender's transmit interval and the receiver's receive interval. Cloud Router's detection time is the peer's agreed transmit interval multiplied by the peer's multiplier, so the peer's intervals must come down too. With default settings BFD detects failure in 5 seconds instead of 60 for BGP.

Question 5 · choose 1

The same custom learned route for 192.168.50.0/24 is applied to two BGP sessions of one Cloud Router: on session A with priority 200 and on session B with priority 100. Both sessions are up. Which path does the VPC network prefer for traffic to 192.168.50.0/24?

  1. ASession A, because a higher priority value means a stronger preference
  2. BBoth sessions equally, because custom learned routes ignore their priority
  3. CWhichever session was created first, because custom learned routes have no priority
  4. DSession B, because lower MED values indicate higher priority
Show the answer and why
  • ASession A, because a higher priority value means a stronger preference

    Incorrect

    Custom learned route priorities are MED values, and lower values win.

  • BBoth sessions equally, because custom learned routes ignore their priority

    Incorrect

    The priority of a custom learned route is treated like a MED value in route selection.

  • CWhichever session was created first, because custom learned routes have no priority

    Incorrect

    Each custom learned route has a configurable priority from 0 to 65,535.

  • DSession B, because lower MED values indicate higher priority

    Correct

    A custom learned route's priority is a MED value from 0 to 65,535, and lower MED values indicate higher priority.

Custom learned routes carry a priority that is a MED value between 0 and 65,535. Cloud Router treats it like a MED value when it selects next hops, and lower MED values indicate higher priority, so the route on session B is preferred while that session is up.

Practice domain 4 →Practice all domains →