Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Professional Cloud Network Engineer practice bank has an additional 2,437 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
Your organization connects an on-premises data center to a Google Cloud VPC network (configured with a custom MTU of 8896 bytes) using an HA VPN gateway and a Cloud Router. After a network expansion, you observe two issues:
DF) bit set sent from Compute Engine VMs to on-premises servers are silently dropped without the VMs reducing packet size.ESTABLISHED, with Cloud Logging showing BGP: Sent NOTIFICATION 6/1 (CEASE/MAX_PREFIXES_REACHED).Which set of remediation steps should you take to restore both the UDP data path and the BGP session?
This option adds an egress ICMP firewall rule on the VPC network and tunes BGP graceful restart and hold timer parameters on the on-premises router.
0.0.0.0/0), so adding an egress ICMP rule does nothing to unblock the returning ingress ICMP messages.CEASE/MAX_PREFIXES_REACHED, where Cloud Router administratively closes the session and will never automatically re-establish peering until prefixes are reduced and the peer is manually cycled.MAX_PREFIXES_REACHED notification.The BGP outage were caused by brief BGP session flaps during routine Cloud Router control-plane maintenance rather than a prefix-limit violation, and an explicit deny-egress firewall rule had been blocking outbound ICMP probes.
This two-part remediation resolves both the data-plane Path MTU Discovery (PMTUD) black hole affecting non-TCP traffic and the control-plane BGP prefix limit shutdown on the Cloud Router.
SYN and SYN-ACK packets, which explains why large TCP transfers succeed automatically. However, UDP has no MSS field and relies on PMTUD. Because Google Cloud does not perform IP fragmentation inside a VPC path, oversized 1,800-byte UDP packets exceeding the VPN tunnel MTU are dropped, and Google Cloud (or an intermediate router) sends an ICMP Fragmentation Needed message back to the sending VM. Because that ICMP packet arrives with a source IP matching the original destination, an ingress allow VPC firewall rule for ICMP is mandatory for the VM operating system to receive the notification and lower its packet size.NOTIFICATION 6/1 (CEASE/MAX_PREFIXES_REACHED), tears down the BGP session, and refuses to re-establish it automatically. Summarizing or filtering on-premises advertisements below 5,000 prefixes and then manually disabling and re-enabling the BGP peer clears the shutdown state and restores dynamic routing.8896-byte jumbo-frame VPC MTU for internal VM-to-VM traffic while allowing heterogeneous MTUs across hybrid VPN links.ESTABLISHED throughout the BGP reset.Unlike changing the VPC MTU globally or tearing down IPsec tunnels, permitting ingress ICMP for PMTUD and cycling the BGP peer after prefix reduction directly targets the exact protocol mechanics causing both failures with minimal disruption.
L3_DEFAULT protocol so Google Cloud fragments oversized packets; filter the on-premises route advertisements below 5,000 prefixes and wait for the on-premises router to automatically re-establish the BGP session.This option routes hybrid UDP traffic through an internal passthrough Network Load Balancer using the L3_DEFAULT protocol to force network-side fragmentation, while filtering routes solely on the on-premises router.
L3_DEFAULT on an internal passthrough Network Load Balancer or protocol forwarding rule allows the forwarding rule to process both the primary protocol (such as UDP) and ICMP PMTUD messages—it does not cause Google Cloud to perform IP fragmentation. Furthermore, the packets have the Don't Fragment (DF) bit set, which prohibits fragmentation anyway, and ingress ICMP must still be allowed by VPC firewall rules.CEASE/MAX_PREFIXES_REACHED event. Waiting for the on-premises router to initiate a connection will fail indefinitely until you explicitly disable and re-enable the BGP peer on the Cloud Router.L3_DEFAULT is specifically used on forwarding rules so PMTUD ICMP messages follow the same load-balancing/forwarding path as the original flow, not to enable packet fragmentation.MAX_PREFIXES_REACHED to protect VPC routing tables from runaway route leaks until an administrator intervenes.An internal passthrough Network Load Balancer with a UDP-only forwarding rule was dropping returning PMTUD ICMP packets (requiring L3_DEFAULT), and the BGP session had merely reset due to a transient link flap rather than hitting the 5,000-prefix limit.
1460 bytes and enable VPC-level TCP MSS clamping for UDP traffic; delete and recreate the HA VPN tunnels to reset the Cloud Router BGP session state.This option lowers the VPC network MTU to the default 1460 bytes, attempts to apply TCP MSS clamping to UDP flows inside the VPC, and tears down the HA VPN tunnels.
SYN and SYN-ACK packets. UDP headers have no MSS field, so MSS clamping cannot resize 1,800-byte UDP datagrams.1460 does not fix tunnel MTU drops: Even with a 1460-byte VPC MTU, an encrypted HA VPN tunnel still has a smaller effective inner MTU (due to IPsec ESP and NAT-T encapsulation overhead), and 1,800-byte UDP packets with the DF bit set will still fail if ingress ICMP for PMTUD remains blocked by VPC firewalls.ESTABLISHED HA VPN tunnel causes unnecessary data-plane outage without fixing the 5,000-prefix limit on Cloud Router; as soon as the on-premises router advertises the same route table, Cloud Router will immediately shut down the BGP session again.The tunnel itself had failed IPsec Phase 1/Phase 2 negotiation due to mismatched IKE profiles, and internal VPC workloads needed a uniform 1460-byte MTU.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated