BENEFITS AND HAZARDS TO ORGANIZATIONS THAT USE SOLID-STATE DRIVE (SSD) STORAGE DEVICES
According to Anderson, Gagliardi, May, McCright, Tlusty & Varela (2009), Solid State Drive (SSD) storage devices have several benefits over traditional hard disk drives (HDDs). Anderson et al (1999) further noted that with no rotational or seek time delays, SSDs have the potential to deliver considerably better I/O performance than HDDs. Pham & Emami (2012) have also pointed out that SSDs deliver reliability, performance, space, as well as energy efficiencies for data centers that run applications with greater input/output operations per second (IOPS) requirements. However, one of the greatest shortcomings of SSD is cost. Additionally, SSDs do not display warning signs before failing, and their reliability or longevity, as well as their long term performance is questionable (Pham & Emami, 2012).
One of the primary advantages of SSDs is that they are faster. Pham & Emami (2012) have noted that HDDs have for long served as the main storage medium for enterprise data centers. Their ability to execute hundreds of IOPS and superb per-gigabyte value for average workloads has earned HDDs a place in the corporate computing environment. However, there is an increasing pressure on IT managers to deliver greater or improved performance for applications that are data intensive, such as data analytics, data warehousing, virtualization, as well as other applications that can over-utilize HDD solutions (Pham & Emami, 2012).
Pham & Emami (2012) have further written that enterprise SSDs within high-performance storage arrays provide unique advantages in both value and performance for applications that need substantially high IOPS. Due to these advantages, organizations are progressively deploying SSD solutions that have the potential to deliver tens of thousands of IOPS per SSD device and use up fewer data center resources (Pham & Emami, 2012).
For instance, Sliwa (2012) has pointed out that Baron Capital Inc. uses single-level cell (SLC) SSDs for the top line of its Dell Compellent Series 3 array primarily to boost the performance of their SQL server. Sliwa (2012) further pointed out that a senior systems administrator at the financial services firm acknowledged that SSDs are particularly helpful when it comes to database transaction logs. According to the senor systems administrator, most of the read/writes occur in the database transaction logs, and placing them on SSD dramatically enhanced the performance of the database (Sliwa, 2012).
Pham & Emami (2012) have also suggested that SSDs offer better value for numerous high-IOPS workloads. Enterprises can substitute as many as 20 HDDs with one SSD that offers greater or equivalent performance in a significantly smaller footprint for data-intensive applications. For a needed level of input/output intensity, there is a peak disk drive solution, either flash or hard disk. The performance intensity requirements will determine the mark at which single-level-cell (SLC) SSDs deliver more value than either Serial Advanced Technology Attachment (SATA) SSD or Serial-attached SCSI (SAS). Due to the fact that more HDDs would be needed to satisfy the higher IOPS (workload) requirements, SSDs offer the better value when the needs hit approximately 120 IOPS/GB.
For instance, Sliwa (2012) wrote that a storage administrator at a major retailer acknowledged that the organization’s IT department bought a couple of SSDs for its Compellent array instead of purchasing 20 or more 15,000 rpm drives after getting information from an Exchange Server expert on the IOPS needs for the transaction logs in the new version of the mail server.
Pham & Emami (2012) have noted the value of SSD in the real world. According to Pham & Emami (2012), flash memory SSDs provide the best value for the applications having high IOPS intensity. Storage and platform vendors utilize flash memory SSDs for resolute storage or posthaste to develop expanded caching devices. Flash memory technology has the ability to assist eliminating I/O bottlenecks and enhance overall performance. For example, NetApp uses an SSD cache to speed up read I/O performance Pham & Emami, 2012).
With SATA and SAS interface, as well as form factor compatibility, SSDs have the capacity to be used with the existing disk back-planes, protocols and storage controllers (Pham & Emami, 2012). SSDs do not have moving parts, electromechanical arms, spinning platters, and motors. SSDs boast of greater mean time between failure (MTBF) with innate resistance to vibration, temperature variances and shock. According to parts stress analysis tests, there is a MTBF of 2 million hours for Lightning Write-Intensive SAS SSDs. On the contrary, HHD vendors can typically claim a mean time between failures (MTBF) of 1.2 million hours. Nevertheless, HDD vendors acknowledge the fact that in the real world setting, this figure might be significantly lower than the actual number (Pham & Emami, 2012).
According to Pham & Emami (2012) , there are several disadvantages or hazards to SSDs. To begin with, the price tag for SSDs per gigabyte is far much greater than HDDs. This means that an upgrade to a similar GB capacity has the potential to incur some significant costs. Another disadvantage of SSDs that has been noted by Pham & Emami (2012) is that while SSDs have the ability to withstand movement, they are vulnerable to loss of power and magnetic or electrical currents much in the same manner as flash cards. Currently, the number of large capacity SSD models is limited. However, this situation is expected to change dramatically over the course of the next few years. Pham & Emami (2012) has also noted that SSDs have a limited write cycle when compared to HDDs. According to the most recent estimates, these write cycles will continue to exist until long after the system is still under use. In fact, there is a possibility that some files could utilize write cycles frequently enough that the owner or user is affected. Moreover, despite the fact that SSDs need less power, many SSDs still use more power than the currently existing standard HDDs. This is particularly true when the device is idle. This can cause the particular device to use up energy at a faster rate (Pham & Emami, 2012).
Anderson, M. Gagliardi, R. May, H. McCright, G. Tlutsy, S. & Varela W. (2009) Performance Value of Solid State Drives using IBM i. Retrieved from: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0ahUKEwjj597r1c_KAhWBAxoKHTGfCjwQFggcMAA&url=http%3A%2F%2Fwww.ibm.com%2Fsystems%2Fresources%2Fssd_ibmi.pdf&v6u=https%3A%2F%2Fs-v6exp1-ds.metric.gstatic.com%2Fgen_204%3Fip%3D126.96.36.199%26ts%3D1454092811842348%26auth%3Dacnmvxexbw3h7gfuvxwsnoqkjfjzv7cs%26rndm%3D0.7033630063481217&v6s=2&v6t=2834&usg=AFQjCNHnLQQZH1AJA972X6O8cgNcUQ7Bcw&bvm=bv.113034660,d.d2s
Pham, D. & Emami, D. (2012) Enterprise SSD Storage Solutions: Engineered for Reliability, Speed, and Predictable Endurance for Maximum Application Value. Retrieved from: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0ahUKEwjfj9Xu18_KAhVMuBQKHSumCYUQFggcMAA&url=http%3A%2F%2Fwww.sandisk.com%2Fassets%2Fdocs%2FNetApp%2520Write-Intensive%2520SSD%2520White%2520Paper.pdf&usg=AFQjCNEpNaGJRlQmV1uUekh1Bm6IraXwDQ&bvm=bv.113034660,d.bGQ
Sliwa, . (2012) “Enterprise IT Shops now choose SSD Storage”. Retrieved from: http://searchsolidstatestorage.techtarget.com/feature/Enterprise-IT-shops-now-choose-SSD-storage