Cloudware Operations and Management - Guide to Cloud Computing for Business and Technology Managers - page 376

18.2.5.5 Authorization and Authentication
In public clouds, safeguards must be placed on machines to ensure proper
authentication and authorization. Within the private cloud environment, one
can track, pinpoint, control, and manage users who try to access machines
with improper credentials. Single sign-on is the basic requirement for a cus-
tomer who accesses multiple cloud services.
18.2.6 Fault Tolerance
In case of failure, there will be a hot backup instance of the application,
which is ready to take over without disruption. Cloud computing outages
extend into the more refined version of cloud service platforms. Some out-
ages have been quite lengthy. For example, Microsoft Azure had an outage
that lasted 22 h on March 13-14, 2008. Cloud reliance can cause significant
problems if downtime and outages are removed from your control. Table 18.1
shows failover records from some cloud service provider systems. These are
significant downtime incidents. Reliance on the cloud can cause real prob-
lems when time is money.
Google has also had numerous difficulties with its Gmail and application
services. These difficulties have generated significant interest in both tradi-
tional media and the blogosphere owing to deep-seated concerns regarding
service reliability. The incidents mentioned here are just the tip of the ice-
berg. Every year, thousands of websites struggle with unexpected down-
time and hundreds of networks break or have other issues. So, the major
problem for cloud computing is how to minimize outage/failover to pro-
vide reliable services. It is important to adopt the well-known Recovery-
Oriented Computing (ROC) paradigm in large data centers. Google uses
Google File System (GFS) or distributed disk storage; every piece of data is
replicated three times. If one machine dies, a master redistributes the data
to a new server.
18.3 Core Portfolio of Functionality
This section discusses the functional components that have to be provided
by the service provider. All the cloud service providers may not provide all
of these components; on the contrary, additional components may be added
if the need arises. The delivery of these functional components may be based
on different optimization criteria like cost optimization or performance
optimization and fulfill different user constrains like budget, performance,
instance types, load balancing, instance prices, and service workload.