Hardware Reference
In-Depth Information
6.5 Cloud Computing: The Return of Utility Computing
If computers of the kind I have advocated become the computers of the future, then computing
may someday be organized as a public utility just as the telephone system is a public utility….
The computer utility could become the basis of a new and important industry.
John McCarthy
MIT centennial celebration (1961)
Driven by the demand of an increasing number of users, Internet companies such as
Amazon, Google, and Microsoft built increasingly larger warehouse-scale computers from
commodity components. This demand led to innovations in systems software to support op-
erating at this scale, including Bigtable, Dynamo, GFS, and MapReduce. It also demanded im-
provement in operational techniques to deliver a service available at least 99.99% of the time
despite component failures and security atacks. Examples of these techniques include fail-
over, firewalls, virtual machines, and protection against distributed denial-of-service attacks.
With the software and expertise providing the ability to scale and increasing customer led
mand that justified the investment, WSCs with 50,000 to 100,000 servers have become com-
monplace in 2011.
With increasing scale came increasing economies of scale. Based on a study in 2006 that com-
pared a WSC with a datacenter with only 1000 servers,
Hamilton [2010]
reported the following
advantages:
■
5.7 times reduction in storage costs
—It cost the WSC $4.6 per GByte per year for disk storage
versus $26 per GByte for the datacenter.
■
7.1 times reduction in administrative costs
—The ratio of servers per administrator was over
1000 for the WSC versus just 140 for the datacenter.
■
7.3 times reduction in networking costs
—Internet bandwidth cost the WSC $13 per Mbit/sec/
month versus $95 for the datacenter. Unsurprisingly, you can negotiate a much beter price
per Mbit/sec if you order 1000 Mbit/sec than if you order 10 Mbit/sec.
Another economy of scale comes during purchasing. The high level of purchasing leads to
volume discount prices on the servers and networking gear. It also allows optimization of the
supply chain. Dell, IBM, and SGI will deliver on new orders in a week to a WSC instead of 4 to
6 months. Short delivery time makes it much easier to grow the utility to match the demand.
Economies of scale also apply to operational costs. From the prior section, we saw that many
datacenters operate with a PUE of 2.0. Large firms can justify hiring mechanical and power
engineers to develop WSCs with lower PUEs, in the range of 1.2 (see
Section 6.7
)
.
Internet services need to be distributed to multiple WSCs for both dependability and to re-
duce latency, especially for international markets. All large firms use multiple WSCs for that
reason. It's much more expensive for individual firms to create multiple, small datacenters
around the world than a single datacenter in the corporate headquarters.
Finally, for the reasons presented in
Section 6.1
,
servers in datacenters tend to be utilized
only 10% to 20% of the time. By making WSCs available to the public, uncorrelated peaks
between different customers can raise average utilization above 50%.
Thus, economies of scale for a WSC offer factors of 5 to 7 for several components of a WSC
plus a few factors of 1.5 to 2 for the entire WSC.
While there are many cloud computing providers, we feature Amazon Web Services (AWS)
in part because of its popularity and in part because of the low level and hence more flexible
Search WWH ::
Custom Search