An administrator Is responsible for managing a VMware Cloud Foundation (VCF) fleet. The administrator discovers intermittent performance issues with the supplemental storage (ISCSI) connected to VCF…

Question

Options

• AReview the iSCSI target's configuration to ensure it's configured for maximum performance,
• BExamine the iSCSI VMkernel port on all affected ESX hosts for TCP retransmissions and
• CUpdate the network plugin on the ESX host to the latest version.
• DEnsure all ESX hosts have the VMkernel port MTU set to 1500.
• EEnsure all ESX hosts have the VMkernel port MTU set to 9000.

• A
  4% (1)
• B
  85% (22)
• C
  8% (2)
• D
  4% (1)

Explanation

To diagnose and resolve the intermittent performance and connection drop issues with the supplemental iSCSI storage, the administrator should focus on network layer consistency and health, particularly regarding packet size (MTU) and delivery (TCP). Examine the iSCSI VMkernel port for TCP retransmissions (Action B - Diagnose): "Intermittent" connection drops and slow I/O are classic symptoms of packet loss or fragmentation issues. By examining the ESXi network stats (e.g., using esxtop key n or viewing vSphere performance charts) for TCP retransmissions, the administrator can confirm if packets are being dropped or lost in transit. Checksum offload errors can also indicate issues where the NIC hardware is incorrectly validating packets, causing the OS to drop them. This step identifies the root cause (packet loss/corruption). Ensure all ESX hosts have the VMkernel port MTU set to 9000 (Action E - Resolve): For high- performance storage traffic like iSCSI in a VMware Cloud Foundation environment, it is best practice to use Jumbo Frames (MTU 9000) end-to-end (Host -> Switch -> Storage Array). The symptom that some hosts are affected suggests configuration drift where those specific hosts might be set to a different MTU (e.g., 1500) or are mismatched with the physical network/target (which is likely set to 9000 for performance). An MTU mismatch (e.g., Target sending 9000-byte frames to a Host/Switch expecting 1500) typically results in the "Do Not Fragment" (DF) bit causing packet drops, leading to the reported connection drops and retransmission delays. Ensuring a consistent MTU of 9000 across the fleet resolves this and aligns with VCF performance standards.

No community discussion yet for this question.