It doesn’t take much to see that the exponential growth of incoming content (data and electronic files) from all sources is overtaking your ability to acquire more storage. To keep your head above water, you need strategies for controlling the growth. You are storing content from many different sources, and although the composition of the content varies dramatically, the strategy is the same. You need to identify what you need to keep, and you need the freedom to destroy what you don’t need to keep.
Simple as that sounds, it can be extraordinarily complex. Considering the volume and diversity of the content, identification can be problematic. Considering the litigation risk and compliance issues, if you can’t confidently identify your important records, you won’t have the freedom to destroy anything. Fortunately, because you Run Better with SAP, the same disciplines and best practices that enable you to structure your master data and processes in SAP applications will come to your rescue when classifying your content.
In my last blog – “Is Your Data Ready for the New Year?” – I wrote about legacy data systems and their impact on your Green IT initiatives. This time, I consider your production SAP database and your strategy for archiving historical data. Because we are talking about structured SAP data, it’s easy to identify what you need to keep and what you can archive. I will cover the more challenging types of content in subsequent posts.
According to the 2008 ASUG Data Archiving Survey, the average SAP database is larger than 1.6 TB and 15% are larger than 6.4 TB. The average database is growing by more than 400 GB per year and 10% are growing by more than 1.5 TB. Most SAP customers eventually take advantage of the data archiving features built into the SAP system – 80% of those surveyed have archived some of their data. The main reasons for archiving are Conversion or Migration, Performance, Hardware and Admin Cost, System Availability, and Legal Requirements.
Your production SAP database is stored on high-performance storage hardware – what I call “Tier 1” storage. This is typically the most expensive type of storage. The archived data is stored in an archive storage system – what I call “Tier 3” storage – in a compressed format, so it occupies only about 20% of the original size. The archive storage is typically the least expensive type of storage.
It is obvious that data archiving reduces data storage requirements. What is not quite as obvious is that the amount of the reduction is amplified because you have multiple copies of your SAP database. You don’t have just a production SAP database. You have additional copies of your SAP database for backup, disaster/recovery, development, testing, and training. Those 3 to 7 copies are most likely stored on Tier 2 storage – a middle class of storage that can be less expensive than Tier 1 storage, but in many organizations is not.
The archive system is sharable, so you don’t need to maintain an archive system for each Tier 2 instance. That means that each terabyte that you eliminate from your production database eliminates 3 to 7 terabytes from your Tier 2 storage.
Let’s take an example of an organization that has been running SAP for several years, has a 1.5 Terabyte SAP database with 5 copies in Tier 2 storage, and has never done any archiving. We’ll assume that the SAP database is growing at a rate of 15% per year, which is well below the average rate of those who responded to the ASUG survey described above. In 10 years, the SAP database will grow to 5.3 terabytes if no data archiving is done.
The table below shows what the SAP database storage looks like before and after implementing data archiving. It also looks out 10 years in the future.
| Tier 1 Storage (gigabytes) (1) | Tier 2 Storage (gigabytes) (2) | Tier 3 Storage (gigabytes) (3) | Storage Cost (4) | Electricity Used (kWh) (5) | Electricity Cost (6) | Carbon Emissions (lbs. CO2) (7) |
|---|---|---|---|---|---|---|---|
Year 1: Before Archiving | 1,500 | 7,500 | 0 | $247,500 | 10,249/yr. | $1,042/yr. | 16,399/yr. |
Year 1: After Archiving (8) | 900 | 4,500 | 120 | $149,940 | 6,286/yr. | $339/yr. | 10,058/yr. |
Year 1 Reduction (Increase) | 600 | 3,000 | (120) | $97,560 | 3,963 | $403 | 6,341 |
Year 10: With No Archiving (9) | 5,277 | 26,384 | 0 | $870,674 | 36,055/yr. | $3,667/yr. | 57,689/yr. |
Year 10: With Archiving (10) | 1,396 | 6,981 | 318 | $234,194 | 266,700/yr. | $2,660/yr. | 41,844/yr. |
Cumulative Reduction (Increase) Over 10 Year Period | 19,135 | 95,677 | (2,128) | $3,131,815 | 128,325 | $13,051 | 205,321 |
Although the first year energy savings are not substantial, the benefits compound year over year. In addition, the SAP database keeps growing year after year, so that the amount of archived data and associated savings increases each year.
What’s more impressive is the money saved from avoided storage cost, which amounts to nearly $100,000 in the first year and accumulates to a total of over $3 million over the 10 year period. This compound benefit effect is found in most Green IT initiatives. So, the sooner that we act, the better off we’ll all be!
Next time I will write about the benefits of removing paper from your organization. Meanwhile, you can learn more about how SAP and its partners support Green IT through the SAP Sustainability Map on the EcoHub.
What do you spend on your SAP data storage? How do you keep data growth under control? Have you considered archiving as a way to reduce storage costs? Do you consider the environmental impact of your data center storage?
References:
(1) Tier 1 storage is the production SAP Database and has High Performance & Availability. This storage is typically the most costly type of storage -- for example, using Fiber Channel.
(2) Assumes 5 copies of the production SAP database.
(3) Assumes archived data is compressed by 80% and shared by all Tier 1 and Tier 2 instances.
(4) According to industry experts, the total cost of storage is comprised of the following component costs: 36% = Hardware (disk); 6% = Software; 44% = Personnel; 1% = Connectivity; 6% = Facilities (environment, physical space, etc.); 6% = Disaster Recovery; 2% = Other (training, etc.). This table assumes that the annual cost per gigabyte is $40 (Tier 1); $25 (Tier 2); and $12 (Tier 3).
(5) Assumes electrical consumption of storage devices is 0.13 kWh per Terabyte. Source: Tape and Disk Costs – THE CLIPPER GROUP Explorer(TM); What It Really Costs to Power the Devices; Analyst: Dianne McAdam; Report #TCG2006046; June 4, 2006
(6) The rolling 12-month average cost of commercial electricity was 10.12 cents per kilowatt hour in August 2010. Source: U.S. Energy Information Agency
(7) The average U.S. CO2 emissions are 1.58 pounds per kWh. Based on 7.18 x 10-4 metric tons CO2 / kWh non-baseload national average emissions rate for converting kilowatt-hours into avoided units of carbon dioxide emissions. Source: eGRID2007 Version 1.1; U.S. annual non-baseload CO2 output emission rate, year 2005 data U.S. Environmental Protection Agency, Washington, DC.
(8) Assumes that 40% of the production data is archived in Year 1.
(9) Assumes that the SAP production database grows at an annual rate of 15%.
(10) Assumes that 10% of the data is archived in each of Years 2-10.