With two controllers, the RAID array generates nice throughput performance. However, the results seem somewhat sobering because, on average, the RAID array delivers a 2245 MB/s read and 2094 MB/s write rate in the fastest configuration. Those are good results, but still a good deal away from the magical 3 GB/s threshold. After breaking down the data according to queue depths, we see a very different picture.
Due to bandwidth limitations of the LSI cards, it really doesn’t matter whether we use the 9280-24i4e with eight or 16 SSDs. We reach the highest read and write rates using the dual-controller solution, driven by two 9260-8i cards, as each is connected to a different PCI Express 2.0 port.
At low queue depths, the different test systems are pretty close to each other. But at a queue depth of four pending commands, the dual-controller system gets going. At QD=8, it really shows its advantage. There, it fluctuates between just below the 3 GB/s mark when writing, and significantly over when reading.
When we connect one controller to our test system, the performance limit is reached much faster. The RAID doesn't exceed 1500 MB/s in writes and around 1600 MB/s in reads.
- The 2011 Extreme RAID Project
- The SSDs: 16 x Samsung 470 (256 GB)
- The Controllers: 1 x LSI MegaRAID 9280-24i4e And 2 x LSI MegaRAID 9260-8i
- RAID Creation In Windows
- Benchmark Results: Throughput
- Benchmark Results: I/O Performance
- Benchmark Results: 4 KB Random Reads/Writes
- Conclusion: Second-Generation 6 Gb/s Systems Needed