The Toshiba MK01GRRB/R is a top-flight enterprise HDD Series that comes in a surprisingly small form factor. Today we are testing the 147GB version, the MK1401GRRB that is in the same product line as the MK3001GRRB/R, which is the 300GB bigger brother. While 300GB might not seem to be a huge amount of capacity, for a 2.5" 15,000 RPM drive, it the highest available at the time of writing.
This small form factor allows for 2.5" HDDs such as the MK1401GRRB to provide more performance in higher density enterprise HDD applications. Increased storage density leads to less space and cooling requirements. This reduced footprint requires less rack space and floor space to provide more performance than 3.5" form factor HDDs.
Every square foot of space is crucial in datacenters, as it has to be cooled, raising the overall power requirement for the data center. One of the greatest aspects in utilizing 2.5" hard drives is the lower wattage requirements than their 3.5" counterparts. Lower wattage results in less heat generation from the device itself. Toshiba has also integrated enhanced idle spin-down power states to reduce power consumption further.
And then we come to performance. Operating at 15,000 RPM places these HDDs at the top of the pack when it comes down to sheer performance. 15K drives provide a great combination of speed, capacity and agility. With highly random workloads becoming more commonplace as virtual machines proliferate in the data center, hard disk drives that can provide increased performance are in high demand.
Protecting data is also very important, and to that end, Toshiba has provided a version of the MK01GRRB series that includes AES-256 device-level encryption without sacrificing performance. This model is denoted with an R at the end of the product SKU instead of a B. This functionality allows customers to repurpose, service, return or retire drives without long overwrite operations or physical destruction of the HDDs when critical data is in place.
At the end of the day getting more performance, higher density, and lower heat generation with less power sounds like a dream come true for datacenter administrators. Supplying a five year warranty provides customers with some peace of mind, so let's take a look and see if the Toshiba MK01GRRB series delivers on these promises.
It wasn't too long ago that many were decrying the 15,000 RPM HDD as dead, with the performance of the SSD being too much for the 15K HDDs to compete with. Unfortunately, for the SSD industry, this hasn't proven to be true. Currently SSDs consist of only 3% of enterprise storage capacity. 15K drives continue to thrive and grow in the datacenter.
Last year less than two million enterprise class SSDs shipped, compared to roughly 56 million enterprise HDDs. The enterprise HDD market is projected to grow 25% in units and 83% in total capacity this year alone. The data deluge has been brought on by the movement of data from the client to the cloud. This is partly due to the explosion of mobile computing, and the datacenter is in dire need of higher capacity storage devices that can also yield great performance.
SSDs will be an important component in the storage industry, but will do little to meet the demand for the Exabyte's of data required. Therefore, SSDs will continue to reside in the Enterprise HDDs shadow in terms of units shipped for the foreseeable future. The reality of the fact is that in the future SSDs and HDDs will both be prevalent in the datacenter, with tiered storage of data becoming more commonplace. Current data tiering strategies focus around three distinct layers, or 'tiers'.
Slower media in the bottom tier typically stores static data that isn't frequently accessed. This storage tier typically consists of tape and slower 7,200 RPM Nearline or SATA 3.5" HDDs. The middle tier of data storage holds more frequently accessed data stored on faster spinning media such as enterprise-class SAS 10,000 RPM HDDs.
The top tier is the 15,000-RPM SAS HDDs. This tier is all about speed of service, with only the most commonly accessed 'hot' data being stored on this more expensive tier of premium data storage. Estimates from many sources indicate that the 'hot' data in the majority of applications comprises less than 10% of the available storage pool.
More software caching solutions are becoming available that can intelligently identify and classify data. These solutions automatically move data between the tiers, optimizing the entire storage pool. The availability of these types of software tiering and caching solutions is growing, especially as SSDs gain more penetration into the datacenter. SSDs will begin to create a fourth tier for the most frequently accessed and modified data.
These tiers aren't rigid, and much of the segmentation will be decided by the type of workload and the application that is applied. Many of the standardized workloads listed above are actually well suited for high-performance HDD tiers. The workloads with high percentages of read activity, paired with low percentages of random writing, are particularly well suited for HDD usage.
The key takeaway is that HDDs are not displaced by SSDs, but merely another high performance layer will be introduced into the picture with SSDs residing in the top tier.
Toshiba MK1401GRRB Internals
The top of the drive reveals the drives small 2.5" form factor with 15mm Z-height. The firmware for the model we are testing is 0102, among the other relevant information listed on the sticker.
The bottom of the drive shows that all of the components on the PCB are inward facing. This is for a number of reasons, from more effective thermal dissipation to also keeping fragile protruding components safe from possible rough handling during installation.
The drive features a standard 6GB/s SAS connection. SAS is much better suited for the enterprise space than SATA. SAS enables many more enterprise features and allows for longer cabling.
Once we remove the PCB, we can see that it lays nearly flush to the bottom of the HDDs metal shell. The PCB is mounted so closely that there is a clear plastic divider between the two. We can also observe the two holes in the plastic separator contain some bare metal with a thermal interface material applied. This TIM allows the Marvell processor and the motor controller onboard the PCB to dissipate the heat into the shell of the HDD effectively. Constructed to very high tolerances, the chips actually make contact with the body of the drive.
The inward facing components on the PCB include a Marvell drive controller, the motor controller and the 32MB DDR2 Winbond cache module. The Marvell and motor controller components also have spots on them where the TIM is placed between them and the shell of the HDD. In effect, the HDD case acts as a large heatsink.
Test System Methodology and Product Specifications
Test System Methodology
We are utilizing a new approach to storage testing here at TweakTown for our Enterprise Test Bench. This new approach will apply to both HDD and SSD testing. The inaugural launch of our new methods will be with this Toshiba HDD, so comparisons to other HDDs will be forthcoming as we test new models and HDDs from different manufacturers.
Our new approach centers on providing results that are outside of the typical average measurements recorded in most test scenarios. The problem with average results is that they do little to indicate the variability experienced during the testing period. Measurements of the maximum latency experienced also do little to measure the total distribution of read and write activity during the test. These results only list the single highest I/O, but do not give us an accurate picture of whether or not these outlying I/O's occur once or a number of times during the test session.
Predictability of service can best be measured and tested as a function of the performance over time. We will provide several measurements of each tested setting to provide a better view of the extended performance of the device. Each value is measured fifty times, with ten second intervals between each measurement. The line that extends between the individual measurements reflects the average speed during the test of the selected variable.
Above we can see an example of the 4K random performance of the Toshiba MK1401GRRB that we are testing today. The top line of results on this scatter chart measures the performance with the write caching enabled on the HDD, and the bottom line of results provides the measurements with the caching of the HDD turned off. This chart actually consists of 800 separate measurements of the HDDs performance.
Typical testing would reveal that the average speed is much higher with the caching enabled, represented in this chart as well. Not revealed in typical testing is the much tighter performance and less variability at the higher QD with write caching enabled. This lower amount of variability speaks volumes to the performance of the solution. Even with two devices with similar average speed, there can be a large difference in overall performance. This is illustrated well by observing how much 'scatter' there is in the returned measurements.
In deployment, many applications can hang or lag as they wait for one I/O to complete. This type of testing illustrates the performance variability expected in these types of scenarios. As we begin to compile more HDDs to provide comparisons, we will be able to look at many different aspects of performance that are not typically measured.
For instance, many would assume that the higher results achieved with caching enabled are merely the result of the write I/O being cached entirely. This is not entirely true, as the results remain the same even over an extended period, even when there is more data written than the cache can hold. The reason that we are experiencing much better performance is probably due to the HDD using the cache as a staging area for the random writes, and then converting them to sequential writes when copied down to the platters. This write combining results in much higher performance from the device when dealing with random workloads, typically the weakness of any HDD. All further tests of the Toshiba in this product evaluation are conducted with write caching enabled.
Toshiba does not release overall performance expectations for their HDDs in terms of IOPS or MB/s, instead relying upon the expected seek latency.
We can see that this is a 6GB/s SAS HDD with a power draw of 4.5 Watts for the 300GB and 4.3 Watts for the 147GB model.
Base Product Specifications
The five major measurements of base performance of any storage solution are latency, random read/write and sequential read/write speed. These are the most common measurements posted by manufacturers to advertise storage performance. We will cover the latency and sequential read and write on this page, and random performance on the following page. We begin with a measurement of the latency of the device. The industry standard for measurement of latency is 4K Random Access at a Queue Depth (QD) of one.
The drive exhibits consistent latency across the different amounts of QD, and scales nicely as we enter the higher QD. The drive averages 4.9ms at a QD1.
The 128K sequential read speeds reflect the maximum sequential throughput of the drive using a realistic file size actually encountered in an enterprise scenario. Here we can see that the drive reads very steadily at an average of 1545 IOPS, or roughly 190MB/s.
The HDD reaches an average of 1498 IOPS, or 185MB/s in 128K sequential writes at QD32. There is a bit of variability here, but the scale of the chart is very low. There is only a separation of 65 IOPS between the highest and lowest results in this test.
4K and 8K Random
4K random read speed measurements are an important metric when comparing drive performance, as the hardest type of file access for any storage solution is small-file random.
The 4K random read IOPS scales very well up through the higher Queue Depths, peaking at QD 64 with a maximum value of 540 IOPS. The average speed at QD64 comes in at 530 IOPS.
The 4k write performance peaks at an average of 485 IOPS at QD32, with the maximum value coming in at 497 IOPS at QD64.
8K random read and write speed is a metric that is not commonly tested for consumer use, but for enterprise environments this is an important aspect of operation. With several different workloads relying heavily upon 8K performance, we include this as a standard with each evaluation. Many of our Server Emulations on the following pages will also test 8K performance with various mixed read/write workloads.
For 8K 100% Random Read the Toshiba delivers from 200 IOPS at QD1 up to and average of 521 IOPS at QD64. The maximum speed measured was 528 IOPS at QD64.
The 8K Random write speed holds very steady, starting at 400 IOPS at QD1, and scaling up to 477 IOPS at QD32, with a top speed of 488 IOPS recorded at the same queue depth.
We use several industry standard profiles configured to emulate the data access patterns utilized by a number of different types of servers. These tests are just as much about the ability of the storage solutions to handle different percentages of read/write ratios at the 8K size as they are about emulating workloads.
This first test emulates Database and On-Line Transaction Processing (OLTP) workloads. OLTP is in essence the processing of transactions such as credit cards used heavily in the financial sector. 15,000 RPM HDDs are well suited for this type of IOPS-heavy workload, outperforming the 10K HDDs easily. Databases are the bread and butter of many enterprise deployments, and their data access patterns are similar to OLTP. These are demanding workloads with 8K random of 66% read and 33% write distribution that can bring even the highest performing solutions down to earth.
The Toshiba peaks at QD64 with an average of 524 IOPS, and a highest recorded speed of 536 IOPS.
This test emulates a typical Email Server with a 50% Read and 50% write distribution of 8K random files. 15K HDDs can provide an enhanced amount of IOPS capability, equating to a higher average CPU utilization of the server, maximizing the efficiency and TCO/ROI of the overall system.
The Toshiba gives a peak speed of 529 IOPS at QD64, and an average of 517 IOPS at QD64.
The File Server profile represents typical workloads encountered in file servers. The inherently random nature of data access in file servers can gain considerable variability in performance from different HDDs, and benefit greatly from caching and tiering solutions. This profile tests across a wide variety of different file sizes simultaneously, with an 80% read and 20% write distribution.
The drive maxes out at an average of 516 IOPS at QD64.
The Web server profile is a read-only test with a wide range of file sizes. Web servers are responsible for generating content for users to view over the internet, much like the very page you are reading here. The speed of the underlying storage system has a massive impact on the speed and responsiveness of the server that is hosting the websites, and thus the end-user experience.
A single Toshiba delivers a top average speed at QD64 of 512MB/s.
The Workstation profile emulates the usage patterns of a storage system on a typical user's workstation. This would be the closest to an actual operating system pattern that a user would experience on their desktop. This test is comprised of 8K random access with an 80% read and 20% write distribution.
The Toshiba generates an average of 560 IOPS at QD128.
Power Measurements and Thermal Monitoring
One of the most overlooked areas in many enterprise evaluations of storage solutions is the power consumption and the amount of heat the unit generates. Heat generation is dealt with via a range of different types of active cooling methods. This constant need to dissipate heat away from the datacenter results in one of the highest ongoing expenses in these environments. Active cooling requires power and lots of it.
For every watt of power consumed in a datacenter there also has to be a redundancy for that power as well. This will provide the datacenter the ability to continue operating during power 'events'. This usually consists of large banks of batteries and generators that can be a very expensive proposition.
Limiting the amount of heat introduced into the datacenter reduces the power needed for climate control and the redundancy costs of that power as well.
Power consumption, of both the device itself and the power needed to deal with any heat generation, sometimes costs more than the purchase of the unit itself over the lifespan of the device. Power and heat generation are significant measurements to take into consideration when making purchasing decisions.
The workload testing for heat generation was conducted at a QD of 64. The results are measured as T-Delta to Ambient. This allows for a higher level of accuracy as it accounts for any small variations in room temperature.
We were also able to use AIDA64 software to monitor the temperatures in real-time, and we placed the drive under varying workloads to ascertain the heat production in certain usages. At idle the unit generated 10C, and the highest recorded values were during random reading and writing, generating 14C. Sequential read monitoring shows the device generated 13C, and sequential write measured 11C.
There is little variation in heat generation during the different workloads, with the largest difference being four degrees Celsius while under load.
Overall keeping a low power threshold is the holy grail of high-performance enterprise storage solutions. For every watt of power consumed, there is also an accompanying increase in heat generated by the device. This creates a vicious cycle of overall power consumption as the additional heat generated must also be cooled.
The Toshiba draws power from the 5V and 12V rail, and consumes very little compared to other larger HDDs. The average draw under a number of random and sequential workloads was between 4 and 6 Watts. The HDD drew just over 10 Watts at startup.
IOPS/Watts is a calculation used to determine the amount of IOPS given per Watt of power consumed. This will become more relevant as we add more enterprise HDDs to the comparison field. The Toshiba MK1401GRRB generated 137 read IOPS at 4k random, 148 IOPS at 4k random write, and 245 IOPS per watt with sequential reading.
The need for storage capacity in the datacenter is multiplying daily. Larger and larger volumes are becoming available to meet the capacity needs, but smaller and faster devices are needed to deliver performance objectives and deliver on density. The size of the datacenter is growing, and with growth comes efficiency challenges that need addressing.
With each generation of datacenters demanding more floor space, it is becoming increasingly harder to deliver on efficiency. With lower power requirements per IOPS than their larger brethren, the 2.5" HDDs can also help to lower the power threshold. Moving from 3.5" HDDs to 2.5" can cut energy usage by up to 1/3, and drastically reduce the power needed for cooling as well.
Through utilizing smaller form factors, more performance can be crammed into smaller spaces. Utilizing faster HDDs such as the Toshiba MK1401GRRB in caching and tiering models allows for a more efficient use of rack space. Optimizing the entire storage pool with strategically placed tiers of performance-oriented HDDs can deliver big for those facing workload challenges.
Increasing the amount of IOPS provided to the server itself, especially in virtualized environments, allows for a much higher utilization of the CPUs. This leads to a higher ROI and lower TCO, with fewer servers needed for the same workloads. This can also help to reduce floor space, as getting more from the currently deployed servers alleviates the need for more infrastructure.
Toshiba stands behind their product with a full five year warranty, providing a promise of reliability. Sporting the 6GB/s SAS specification provides enterprise-class features, and allows for longer cabling. A 32MB cache buffer provides consistent performance, and helps alleviate performance hills and valleys.
2.5"HDDs are available at prices that are up to eight times lower than comparable capacity enterprise SSDs, which keeps them competitive. The MK01GRRB series of HDDs are at several retailers at reasonable pricing.
Utilization of the MK1401GRRB in RAID arrays can provide high performance in a dense package. We are currently testing eight Toshiba MK1401 HDDs to be utilized in a series of articles that explore the performance of different enterprise software caching and tiering solutions. We will also be posting a product evaluation with the MK1401GRRB's in different RAID configurations in the coming days.
The 15K 2.5" HDD sector is thriving, more so than many expected with the advent of SSDs and their introduction into the enterprise. The future holds these two solutions as complimentary to each other. As the need for datacenter storage continues to grow, 2.5" HDDs such as the Toshiba MK1401GRRB will be there to face the challenges, packing a big punch in a small package.
PRICING: You can find products similar to this one for sale below.
United States: Find other tech and computer products like this over at Amazon.com
United Kingdom: Find other tech and computer products like this over at Amazon.co.uk
Australia: Find other tech and computer products like this over at Amazon.com.au
Canada: Find other tech and computer products like this over at Amazon.ca
Deutschland: Finde andere Technik- und Computerprodukte wie dieses auf Amazon.de