LSI MegaRAID CacheCade Pro 2.0 Review – Exploring TCO

TCO is an acronym which stands for ‘Total Cost of Ownership’. This is one of the most important measurements that must be taken into consideration when deciding whether or not to upgrade. Basically put…What is this going to cost and is it a viable investment?

When confronted with the thought of upgrading to SSDs, many will balk at the prohibitive start up prices. Yes, SSDs are premium storage devices, and they are priced accordingly. One must take into account the price per transaction of the competing solutions. In the end, that is what it all boils down to. When users factor in the cost of purchasing the equipment, the power required for said equipment, and the cooling and space required for everyday operation, then they will be able to calculate the total cost. Divide that by the total number of transactions that can be achieved and that is the base price per transaction.

This is where SSDs can be deceptive to those that aren’t used to the comparative speeds of the two solutions. SSDs can provide up to 1000 times more transactions per second than their HDD counterparts, yet they certainly aren’t 1000 times more expensive!

For those looking to accelerate their already existing infrastructure there are a few options, some which have noticeable drawbacks. As you can see below, adding cache is a common short term solution, leveraging volatile DRAM. There are a few drawbacks to this approach, one being that it is a very expensive upgrade. Another is that, quite simply, there are only so many DIMM slots.  You can only upgrade your memory once, and after that you cannot scale the application of cache further. Data stored in DRAM is also vulnerable to power failure, or even DIMM failure, unless some other method is used to preserve or replicate the data stored in the DRAM.

With the CacheCade Pro solution, the price is 4 times cheaper, yet also allows you to tailor your caching capacity to the needs of the server and use it for multiple HDD volumes. DRAM Read/Writes to DDR pages, and those pages are marked dirty and flushed to non-volatile storage at checkpoints which is a fixed value that does not take into consideration the loading or access performance at the given flush times. Add to that the loss of data integrity due to possible system crashes or errors, and it clearly cannot compete with a CacheCade Pro application and its many steps taken to ensure data protection.

Here is another great graphic that illustrates the cost differences of taking another route and installing a large amount of short-stroked HDDs to boost the speed of the array. A negative aspect of this is, of course, cost. The price of cooling and power will get expensive in the long run. Adding short stroked HDD with excess unused capacity is also an inefficient use of rack space. An astounding statistic from this comparison is that it will take 192 SAS HDD to achieve what 2 SSDs and 14 drives can easily accomplish. The space savings here alone is significant but not easily factored into the comparisons. Cost savings accumulate quickly without the need to supply an enclosure, and power and cooling for such a large array of SAS drives.

blankSAS HDD can also be cost-prohibitive to some users. SATA drives used in conjunction with a CacheCade Pro Virtual Drive can benefit greatly, and easily outstrip the speeds that even an SAS HDD array can achieve under normal circumstances.

All of these numbers only matter if the solution works, so let’s delve into some testing, and see exactly what this solution has to offer.

NEXT: Test Bench and Protocol

~ Introduction ~ Basic Concepts and Application ~

~ Enter Write Caching ~ Exploring TCO ~ Test Bench and Protocol ~

~ Single Zone Results ~ Overlapped Region Results ~

~ Real World Results and Conclusion ~

Leave a Reply

Your email address will not be published. Required fields are marked *