First, the lead time for server provisioning has been reduced by more than
99 percent. When the business requires a new test or development environment,
the IT team can deploy a new Linux virtual server within 20 minutes. There is no
longer any need to source a new physical server, wait for delivery, then install and
configure it—a process which could take six to eight weeks.
“Since the move to Linux on System z, we’ve received some really positive feed-
back from the business,” comments Ted Mansk. “Users really appreciate the fact
that when they need something, we can now get to work on it at once, instead of
having to wait weeks for new hardware to arrive. Equally, because the System z
platform is so much more reliable than our previous infrastructure, we no longer get
any complaints about performance and availability issues.”
Availability and disaster recovery
In fact, since the new infrastructure went into production, BCBSM has not experi-
enced a single incidence of unplanned downtime or underperformance. Equally,
the company’s disaster recovery capability has been improved dramatically by the
new solution.
“With the old distributed architecture, we would have struggled to restore our appli-
cation servers within 48 hours,” says Ted Mansk. “With all these servers running on
System z, we can perform a full disaster recovery at our secondary data center
within 90 minutes—an improvement of nearly 97 percent. In addition, the ease of
maintenance also contributes to improving availability.”
With Linux on IBM System z, BCBSM can achieve near-continuous availability by
reducing the need for planned downtime. It is possible to perform nearly all mainte-
nance to the hardware, z/OS and DB2 while all systems are up and running—a
feature unparalleled in the industry and a specific design goal of the platform.
Realizing the cost savings
Finally, BCBSM expects the new infrastructure to deliver cost savings over and
above the predicted TCO savings in the initial cost-benefit analysis.
“When we performed the initial cost-benefit calculation, we did not factor in the
maintenance and support costs, or the cost of power, cooling and server room
space,” comments Ted Mansk. “The savings in these areas are likely to be consid-
erable. For example, we only need 1.5 full-time employees to manage the entire
System z Linux environment—which would be unthinkable if they were looking after
140 physical servers.”
Solution Components
Hardware
●
IBM System z9® EC
●
IBM System z10™ EC
Software
●
IBM DB2®
●
IBM z/OS®
●
IBM z/VM®
●
SUSE Linux® Enterprise Server
