HP Integrity Superdome 16-way Server A6113A#801 プリント

製品コード
A6113A#801
ページ / 4
Since implementing Superdome servers, reliable, 
remarkable performance has defined the ERP 
environment. After the recent upgrade to Superdome 
servers with Itanium processors, the IT staff saw further 
performance improvements and 25% lower Oracle 
licensing costs. 
Along with moving to the latest Superdome servers, 
the staff transitioned from v2 to v3 of HP-UX 11i 
to capture performance, reliability, and stability 
improvements. “It was part of our strategy to adopt 
the latest supported operating system as we had 
future expansion of the business in mind. HP-UX 11i v3 
showed a performance improvement for Oracle,” says 
S. K. Rajavelu, team lead—UNIX and storage.
He notes that HP-UX’s security features are effective. 
OpenSSL (open source implementation of the 
Secure Socket Layer) mitigates unauthorized access. 
Restriction enforcement by the operating system kernel 
greatly reduces the performance overhead incurred 
with the audit subsystem. Additionally, HP Logical 
Volume Manager (LVM) and Mirrordisk/UX are also 
proving useful. “They provide value-added operational 
support for our environment,” Rajavelu notes.
An improved, global business platform
The staff also notes that the Superdome’s flexible 
capacity, via HP hard partitions (nPars) and virtual 
partitions (vPars), provides the means to consolidate 
and scale up applications. “One of the most 
important features of our Superdome environment is 
virtualization, which improves utilization and helps 
us get the most from our investment,” says Srinivas 
Kalayan, senior infrastructure architect.
True mission-critical systems like the ERP system  
and Oracle Database instances running Dubai  
World require continuous availability and system 
protection. Integral in Dubai World’s disaster  
recovery strategy are two other HP VSE components, 
SGeRAC and Metrocluster. Addressing the availability  
and manageability gaps in an Oracle RAC (Real 
Application Clusters) cluster, SGeRAC provides 
high availability and disaster recovery for the entire 
environment in a single solution, while supporting 
multiple storage deployment options that improve 
manageability. 
Metrocluster extends Serviceguard’s coverage by 
enabling automatic, bi-directional failover of Oracle 
RAC between the primary Port Rashid data center and 
the Jabel Ali disaster recovery site. The two duplicated 
facilities approximately 25 miles (40 km) apart link by 
fiber-optic cable. Metrocluster keeps the databases 
functioning well even during extreme conditions, and it 
enables the RAC database instances to spread across 
the two data centers without any performance impact 
due to inter-site latencies.
“We were concerned about performance impacts 
after a failover, so we required a disaster recovery 
setup to handle site-level failover requirements.  
HP Metrocluster and Serviceguard Extension for RAC 
integrated with virtualization capabilities all work 
together to help us maintain data integrity, flexibility 
and deliver the services our customers expect,”  
says Nair. 
For Oracle ERP and HP software running on other 
HP Integrity servers, the staff leverages several other 
HP VSE components, including Instant Capacity 
(iCAP), Process Resource Manager (PRM), Workload 
Manager (WLM), and Serviceguard solutions.
“An HP VSE definitely increases performance, 
minimizes space, lowers costs by reducing the number 
of physical servers we would need, and makes it 
easier to provide services to the business,” Nair adds.
A secure storage foundation
Metrocluster and SGeRAC integrate with  
HP StorageWorks XP Continuous Access software 
and HP StorageWorks XP12000 Disk Arrays, which 
provide data replication and storage capacity. 
On the two clusters, HP Services deployed  
HP Serviceguard Storage Management Suite to 
prevent lengthy downtime when expanding storage 
space and make the storage pools easier to manage. 
All of these technologies help maintain data access 
should a mishap occur with the primary facility.
“Three years ago we implemented RAC with the 
Storage Management Suite and we haven’t had 
any issues. The HP disaster tolerance solutions work 
well with our robust XP Disk Arrays. The arrays scale 
well, and the cache feature helps with fast reads and 
performance. These products are helping us achieve 
our business continuity requirements and smooth 
operations,” says Rajavelu.
3