HP ProLiant 3000 Compaq Servers: Enterprise Class Performance Leading the Way - Page 5

Integrated Management Solutions

Page 5 highlights

ECG078/0498 WHITE PAPER COMPAQ SERVERS: ENTERPRISE CLASS SYSTEMS 5... ProLiant server models, in both matched-pair and mixed-pair configurations, and supports both Uniprocessor-Uniprocessor and SMP-SMP combinations. The Series F also offers a choice of either Ethernet or ServerNet as the server-to-server interconnect, and provides Compaq's industry standard fibre channel as the interconnect between servers and storage. There are currently over 14 certified configurations available. This number will expand as new servers and combinations are introduced. Compaq is the first company to receive certification on Microsoft's NT Cluster Server with a storage system based on the Fibre Channel technology standard. The ProLiant Cluster/"Series S" solutions are also based on MSCS and support a broad range of ProLiant servers. S Series clusters are characterized by the use of SCSI technology as the server to storage interconnect, and support Ethernet as the server-to-server interconnect. ♦ Level 4 of Compaq's availability involves scalability clusters that deliver increased performance and high availability. Each node in a scalability cluster is active on a separate "copy" of the same application. Since more than one node works on the same application, more computer resources can be applied to the same application, thus increasing performance. Also, since separate copies of the same application reside on all nodes in the cluster, a failure or crash of one node in the cluster will not cause an application outage. Recovery, therefore, is very rapid. Performance Scalability clusters are available now from Compaq for Unix, including Tandem Unix, and will be available in 1999 for NT. ♦ Level 5 HA solutions, also known as Geographically Dispersed Clusters (GDC), are single clusters, a collection of servers connected by a very high speed interconnect and sharing a single heartbeat. In GDCs, the nodes of a cluster are distributed not merely building to building, as in a campus cluster, but city to city, or state to state, or region to region. In the event of failure, the automatic failover will then occur in not just node to node in the same room or building to building, but also city to city, state to state, and region to region, automatically returning compute services back to end users and allowing them to quickly return to full production work. This level of availability solutions will be available in the year 2000, if market demand develops. Software Reliability A recent customer satisfaction survey demonstrated that Compaq servers provide significantly higher software reliability than its closest competitors. Competitive servers were twice as likely to fail as a result of NOS or application software than Compaq servers.1 This is not surprising given Compaq's extensive work with partners such as Microsoft, SCO, Novell, and others to ensure integration excellence of their software on Compaq platforms. Integrated Management Solutions In addition to unprecedented availability and scalability, servers also require integrated management solutions. Compaq delivers the right combination of management and integration tools necessary for establishing and maintaining optimized, consistent systems throughout a customer's distributed enterprise. Compaq's systems management strategy has four major components: • Engineering manageability built into all Compaq products • Advanced automation for integration and management • Partnering with the leaders in systems management to enable broad management of all Compaq products 1 Survey conducted by Research International, December 1997 (margin of error is plus or minus 10%).

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10

W
HITE
P
APER
C
OMPAQ
S
ERVERS
:
E
NTERPRISE
C
LASS
S
YSTEMS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
ECG078/0498
5
ProLiant server models, in both matched-pair and mixed-pair configurations, and supports
both Uniprocessor-Uniprocessor and SMP-SMP combinations.
The Series F also offers a
choice of either Ethernet or ServerNet as the server-to-server interconnect, and provides
Compaq’s industry standard fibre channel as the interconnect between servers and storage.
There are currently over 14 certified configurations available.
This number will expand as
new servers and combinations are introduced.
Compaq is the first company to receive
certification on Microsoft’s NT Cluster Server with a storage system based on the Fibre
Channel technology standard.
The ProLiant Cluster/"Series S" solutions are also based on MSCS and support a broad range
of ProLiant servers.
S Series clusters are characterized by the use of SCSI technology as the
server to storage interconnect, and support Ethernet as the server-to-server interconnect.
Level 4
of Compaq's availability involves scalability clusters that deliver increased
performance and high availability.
Each node in a scalability cluster is active on a separate
"copy" of the same application.
Since more than one node works on the same application,
more computer resources can be applied to the same application, thus increasing
performance.
Also, since separate copies of the same application reside on all nodes in the
cluster, a failure or crash of one node in the cluster will not cause an application outage.
Recovery, therefore, is very rapid.
Performance Scalability clusters are available now from
Compaq for Unix, including Tandem Unix, and will be available in 1999 for NT.
Level 5
HA solutions, also known as Geographically Dispersed Clusters (GDC), are single
clusters, a collection of servers connected by a very high speed interconnect and sharing a
single heartbeat.
In GDCs, the nodes of a cluster are distributed not merely building to
building, as in a campus cluster, but city to city, or state to state, or region to region.
In the
event of failure, the automatic failover will then occur in not just node to node in the same
room or building to building, but also city to city, state to state, and region to region,
automatically returning compute services back to end users and allowing them to quickly
return to full production work.
This level of availability solutions will be available in the
year 2000, if market demand develops.
Software Reliability
A recent customer satisfaction survey demonstrated that Compaq servers provide significantly
higher software reliability than its closest competitors.
Competitive servers were twice as likely to
fail as a result of NOS or application software than Compaq servers.
1
This is not surprising given
Compaq’s extensive work with partners such as Microsoft, SCO, Novell, and others to ensure
integration excellence of their software on Compaq platforms.
Integrated Management Solutions
In addition to unprecedented availability and scalability, servers also require integrated
management solutions.
Compaq delivers the right combination of management and integration
tools necessary for establishing and maintaining optimized, consistent systems throughout a
customer’s distributed enterprise.
Compaq’s systems management strategy has four major
components:
Engineering manageability built into all Compaq products
Advanced automation for integration and management
Partnering with the leaders in systems management to enable broad management of all
Compaq products
1
Survey conducted by Research International, December 1997 (margin of error is plus or minus 10%).