Home, Cost-Based Optimizer, Benchmarks, Server Systems, Systems Architecture, Processors, Storage,
  Storage Overview, System View of Storage, SQL Server View of Storage, File Layout,
  PCI-E, SAS, FC, HDD, SSD Technology, RAID Controllers, Direct-Attach,
  SAN, Dell MD3200, CLARiiON AX4, CX4, VNX, V-Max, HP P2000, EVA, P9000/VSP, Hitachi AMS
  SSD products: SATA SSDs, PCI-E SSDs , Fusion iO , other SSD

I have not been keeping up with the HP products. I have previously commented that I thought the main HP website was absolutely horrible. There was far too much visually stunning images, and it was far too difficult to find the product I was interested in. It seems as if they hired a website designer with deeply devoted to his/her artistic skills and little consideration for the purpose of the website.

Apparently HP has revamped their website with cleaner lines. But I am not planning on using their portal to navigate. Instead, I will use Google or Bing to find the specific product to go directly to that page.


Update 2013-04

Some over the last year or so, HP updated the performance marks for the P2000 G3 MSA. Note: I am expecting this product to be updated in the next year with SAS 12Gbps, but if you must buy now, then do so. There are also now specifications for performance on 1GbE iSCSI, that is, over multiple channels.

2000 Array PerformanceP2000 G3
P2000 G3
P2000 G3
Protocol (host connect)8 Gb Fibre Channel6 Gb SAS10Gb Ethernet
2000 RAID 10 Performance Results
Sequential Reads MB/s1,650 1,650 1,600
Sequential Writes MB/s850850800
Random Mix IOPs 60/40 read/write20,500 23,500 18,500
2000 RAID 5 Performance Results
Sequential Reads MB/s1,650 1,650 1,600
Sequential Writes MB/s1,3001,3501,000
Random Mix IOPs 60/40 read/write14,000 16,000 12,500
2000 RAID 6 Performance Results
Sequential Reads MB/s1,6501,6501,600
Sequential Writes MB/s1,3001,1001,000
Random Mix IOPs 60/40 read/write8,3009,8007,500
Refer to the paper titled "Upgrading the HP StorageWorks MSA2000 (G1 or G2) to the P2000 G3 MSA ", available in the Resource Library at: www.hp.com/go/p2000.

Sequential Read MB/s is up from 1,572MB/s to 1,650MB/s. Sequential Write MB/s is up from 790MB/s to 850MB/s. Max iSCSI 1GbE bandwidth is 550MB/s, presumably over 8 links? P2000 quick specs I am wondering if the reason for the 1,650MB/s peak performance is that there is a PCI-E gen2 x4 bus inside the P2000 G3 that all IO traffic between host and backend must traverse. The net realizable bandwidth of PCI-E gen2 x4 is course 1,600MB/s (well possibly 1,650).

Original P2000 Notes

The P2000 is the current HP entry SAN. The system can be configured with one or two controllers. The back-end to disk drives (and SSD) is 6Gbps SAS over 1 x4 (per controller?) with a 1 x4 SAS expansion port. The controller options are 8Gbps FC (two ports per controller), FC/iSCSI Combo (1GbE), 6Gbps SAS (four ports per controller) and 10GbE iSCSI (two ports per controller).

There are two P2000 chassis models, one for 12 LFF and another for 24 SFF disks. The chassis can accommodate any of the controllers.


The FC model rear view shows 2 controllers.
SAS model, there are 4 SAS ports on the host side, and 1 expansion SAS per controller.

Expansion can be up to 96 LFF (P2000 + 7 D2600) or 149 SFF (P2000 + 5 D2700) drives.


Below are HP's performance results with prelimary for 10GbE iSCSI.

Performance numbers are a guideline as established by tests using RAW I/O in an Operating System Agnostic test lab environment. 144 GB 15K SAS drives were used in a dual controller configuration of 12 vdisks consisting of twelve disks per vdisk, 1.6 TB volumes, and 3 volumes per host. 4 hosts directly attached to the P2000 G3 array were used in this test configuration (results cannot be expected with a single host). Results were achieved in Sequential Writes with 256K blocks; all random tests were based on 8K block sizes.
NOTE: Number and type of applications, drive type and number of drives, operating system used, and the number of hosts will affect overall performance. This table is provided strictly as a test-lab comparison. Note: These numbers reflect a full array configuration with the maximum number of front-end ports, disks, and controllers. The test results shown for the P2000 G3 are preliminary and designed to give a reference point for comparisons. They will be reposted shortly when finalized.

2000 Array PerformanceP2000 G3
P2000 G3
P2000 G3
Protocol (host connect)8 Gb Fibre Channel6 Gb SAS10Gb Ethernet
2000 RAID 10 Performance Results
Sequential Reads MB/s1,572 1,572 1,500
Sequential Writes MB/s790787775
Random Mix IOPs 60/40 read/write18,698 19,682 15,500
2000 RAID 5 Performance Results
Sequential Reads MB/s1,572 1,572 1,500
Sequential Writes MB/s1,1951,1721,000
Random Mix IOPs 60/40 read/write12,511 12,835 10,500
2000 RAID 6 Performance Results
Sequential Reads MB/s1,5721,5721,500
Sequential Writes MB/s1,1731,0241,000
Random Mix IOPs 60/40 read/write7,3628,0446,500
Refer to the paper titled "Upgrading the HP StorageWorks MSA2000 (G1 or G2) to the P2000 G3 MSA ", available in the Resource Library at: www.hp.com/go/p2000.

The random IO valure appear to be about right. In RAID 10, the 60/40 read/write mix result of just under 20K IOPS corresponds to 12K read IOPS and the 8K write IOPS generate 16K IOPS to disk for a 28K IOPS to disk, just about 200 IOPS per disk based on 144 disks.

The sequential performance, 1,572MB/sec on read, is disappointing. The works out to 11MB/sec per disk. Granted the configuration was for maximum random IOPS, and presumably the 1,572MB/sec could have been reached with fewer disks. Still, this looks to be a system limit.

There are 2 x4 SAS 6Gbps back-end port, each capable of 2GB/sec per sec, so 4GB/sec from the SAS side seems possible. The front-end of 4 x 8Gbps FC should be able support 700MB/sec per FC port for 2.8GB/sec. The SAS controllers have 4 ports per controller, so this is clearly for connectivity and not aggregate bandwidth. My suspicion is that the processor in the P2000 is far too weak.

It is really too bad we do not have the powerful Intel Nehalem or Westmere cores in the entry storage system. There is version of Nehalem quad-core with integrated PCI-E (Jasper Forest), but I think the Westmere dual-core (for desktops and mobile) with the 5520 server IOH would make for a powerful, yet reasonably low cost combination.

MS2000 G2 Performance

HP initially put out set of performance results for their previous generation MSA2000 G2. The FC model performance numbers looked right, but the SAS results were ridiculously low. There was a different set of results that showed reasonable numbers for the SAS model. Then HP later updated the performance report on all 3 MSA2000 models together, as shown below.

MSA2 G2 PerformanceMSA2300fc G2MSA2000sa G2MSA2000i G2
Protocol (host connect)4 Gb Fibre Channel3 Gb SAS1GbE Ethernet
MSA2000 RAID 10 Performance Results
Random Reads IOPs22,87421,86113,658
Random Writes IOPs15,00814,49114,399
Random Mix IOPs 60/40 read/write17,87817,70013,493
Sequential Reads MB1,2381,050274
Sequential Writes MBs532532266
MSA2000 RAID 5 Performance Results
Random Reads IOPs22,04421,00012,335
Random Writes IOPs2,7142,7142,714
Random Mix IOPs 60/40 read/write5,3254,9265,325
Sequential Reads MBs1,2381,050274
Sequential Writes MBs725617266
MSA2000 RAID 6 Performance Results
Random Reads IOPs21,97521,00012,292
Random Writes IOPs1,8761,8761,876
Random Mix IOPs 60/40 read/write4,2933,8653,851
Sequential Reads MBs1,2381,050274
Sequential Writes MBs772729266

The sequential bandwidth on the G3 is up by 30-50% from 1,238MB/s on G2 FC and 1,050 on the SAS to 1,572MB/s on the G3. The random mix IOPS is up slightly for RAID 10 and up substantially for RAID 5 and 6. The G3 RAID 5 random mix IOPS numbers look about where they should be, and the G2 RAID 5 is too low, probably indicating a problem with the RAID algorithm.

The original results with poor SAS performance:

Workload MSA2300fcMSA2000saMSA2000i
Host Connect4 Gb Fibre Channel3 Gb SAS1GbE Ethernet
MSA2000 RAID 10 Performance Results (rounded)
Random Reads IOPs21,40010,6008,200
Random Writes IOPs8,8004,9004,500
Random Mix IOPs 60/40 read/write13,5006,8006,100
Sequential Reads MB1,300700300
Sequential Writes MBs560350260
MSA2000 RAID 5 Performance Results (rounded)
Random Reads IOPs20,50010,2007,800
Random Writes IOPs2,4002,0001,600
Sequential Reads MBs1,300700300
Sequential Writes MBs780380270
Random Mix IOPs 60/40 read/write5,9003,3003,200
MSA2000 RAID 6 Performance Results (rounded)
Random Reads IOPs20,50010,1007,800
Random Writes IOPs1,6001,4001,200
Sequential Reads MBs1,300700300
Sequential Writes MBs820380270
Random Mix IOPs 60/40 read/write4,5002,7502,560

Below are the early SAS results that did not agree with the low SAS results above.

WorkloadMSA2300sa G2
Host Connect3Gb SAS
MSA2000 RAID 10 Performance Results (rounded)
Random Reads IOPs21861
Random Writes IOPs14491
Random Mix IOPs 60/40 read/write17700
Sequential Reads MB1049
Sequential Writes MBs531
MSA2000 RAID 5 Performance Results (rounded)
Random Reads IOPs21019
Random Writes IOPs2705
Sequential Reads MBs1051
Sequential Writes MBs644
Random Mix IOPs
60/40 read/write
MSA2000 RAID 6 Performance Results (rounded)
Random Reads IOPs21029
Random Writes IOPs1918
Sequential Reads MBs1054
Sequential Writes MBs736
Random Mix IOPs 60/40 read/write4872

HP StorageWorks 2000 Modular Storage Array (Obsolete)

First, the model name and numbering system for the HP entry storage line is utterly incomprehensible. Perhaps the product manager may have been on powerful medications at the time, or there were 2 PMs who did not talk to each other. The official name seems to be StorageWorks 2000 Modular Storage Array, but the common name seems to be MSA2000 G2 (for the second generation). This name might just apply to the parent chassis family, comprised of the 2012 12-bay enclosure for 3.5in (LFF) drives and the 2024 24-bay for 2.5in (SFF) drives. The controller itself appears to be the MSA2300 with suffix for the front-end interface. There are two models of interest for database systems, the 4Gb/s fiber channel fc model and the 3Gb/s SAS sa model. Do not even think of putting a critical database server on iSCSI. The choice is between fc and sa on the front-end interface. The configured unit might be the 2312 or 2324.

Apparently there is also the StorageWorks P2000 G3 MSA. This appears to consolidate the G2 fc and i (iSCSI) models, with 8Gb/s FC. Above this, HP has the P4000 series. I am not sure how this relates to the EVA 4400 series.

The back-end interface is SAS, and allows both SAS and SATA drives. The back-end can also connect to either additional 12-bay LFF enclosures (MSA2000) or 25-bay SFF enclosures (MSA70). There is the option of having a single controller or dual-controllers. The storage expansion enclosures can have single or dual IO interfaces. My opinion is that SAS for the back-end interface is the right choice. FC incurs a large cost premium and has no real advantages over SAS. A single 4Gb/s FC port has one-third the bandwidth of a 3Gb/s x4 SAS port, and the same ration for 8Gb/s FC to 6Gb/s x4 SAS.

There are 2 FC ports per controller on the fc model, and four SAS ports on the sa model. There is a single (3Gb/s x4?) SAS port on the backend. HP initially put out a performance report showing reasonable performance numbers for the MSA2000 G2 with 96 15K drives on the fc model of 22,800 random Read IOPS and 1,200MB/sec sequential in RAID 10, but ridiculously low numbers of 10,800 IOPS and 700MB/s for the sa model. Either this was a benchmarking mistake, which seems unlikely for HP's history in this area, or there were bugs in the sa software stack. This was later corrected to 21,800 IOPS and 1,000MB/s. This configuration is essentially the maximum for the MSA2000 with 2.5in. The random reads works out to just over 225 IOPS per disk, but the sequential is 12.5MB/sec per disk. I am presuming that 1GB/sec sequential could have been reach with about 40 disks. The Microsoft Fast Track Data Warehouse Reference Architecture 2.0 document seems to indicate that 100MB/sec per disk is possible for 2-disk RAID-1 groups.

See HP MSA2000 Technical Cook Book. for additional details. If the URL is not correct, search either the title or document number 4AA2-5505ENW.