400-007 Question #119: Real Exam Question with Answer & Explanation

The correct answer is A: Change the protocol to CIFS.. The root cause is a speed mismatch: a 10 GbE storage system is sending block-level iSCSI data to a 1 GbE server faster than the server's link can absorb it, causing output queue drops at the switch port. Switching to CIFS (A) is the most cost-effective solution because CIFS (SMB-

Question

You are using iSCSI to transfer files between a 10 Gigabit Ethernet storage system and a 1 Gigabit Ethernet server. The performance is only approximately 700 Mbps and output drops are occurring on the server switch port. Which action will improve performance in a cost-effective manner?

Options

AChange the protocol to CIFS.
BIncrease the queue to at least 1 GB
CUse a WRED random drop policy
DEnable the TCP Nagle algorithm on the receiver

Explanation

The root cause is a speed mismatch: a 10 GbE storage system is sending block-level iSCSI data to a 1 GbE server faster than the server's link can absorb it, causing output queue drops at the switch port. Switching to CIFS (A) is the most cost-effective solution because CIFS (SMB-based file protocol) operates at the file level and includes application-layer flow control and request/response pacing, which naturally throttles the data rate to match what the receiver can process - eliminating the burst-induced drops. Increasing the queue to 1 GB (B) would only delay drops without resolving the underlying rate mismatch and is expensive. WRED (C) actively drops packets to signal TCP congestion but does not solve the fundamental bandwidth bottleneck. The TCP Nagle algorithm (D) coalesces small writes into larger segments, which actually increases latency and is counterproductive for large sequential storage transfers.

Community Discussion

No community discussion yet for this question.

Full 400-007 Practice