HP-UX and Linux Serviceguard Cluster Manageability, June 2007

• HP Serviceguard for Linux and Red Hat Global File System (GFS) provide continuous access to both
applications and data in Linux environments as well as the ability to scale without creating
complexity. Coexistence of HP Serviceguard for Linux and Red Hat GFS has been tested and
validated and is based on a proven high-availability solution and an open source, POSIX-compliant
enterprise cluster file system.
• HP Serviceguard toolkits make it easy to integrate applications into an HP Serviceguard cluster,
providing sample package control scripts to start, stop, and monitor applications running in the
cluster. These high-availability toolkits include:
– HP Serviceguard for HP-UX—Oracle, Apache, Tomcat, and CIFS/9000 – HP Serviceguard for
Linux—Apache, NFS, MySQL, Oracle, Samba, PostgreSQL, Tomcat, and SendMail
– HP Serviceguard Developers Toolbox Program (HP-UX and Linux)—for customers and partners to
facilitate the quick and easy integration of an application with HP Serviceguard on HP-UX and
Linux
The following solutions build on Serviceguard disaster-tolerant capabilities, allowing it to span across
any distance.
Disaster-tolerant solutions for HP-UX
HP offers the following disaster-tolerant solutions for the HP 9000 and HP Integrity server families:
• HP Extended Distance Cluster—An HP Extended Distance Cluster (XDC) has a typical local cluster
design. It protects two geographically dispersed data centers (up to 100 kilometers apart) against
unforeseen downtime, enhancing data and application availability. XDC uses host-based
Mirrordisk/UX for data replication.
• HP Extended Cluster for RAC—HP Extended Cluster for RAC provides the same functionality as the
Extended Distance Cluster, but it is tailored to Oracle RAC environments. Extended Cluster for RAC
delivers failover capabilities for a single RAC database that is replicated and synchronized across
two remotely located data centers (up to 100 kilometers apart). This active–active disaster-tolerant
solution maintains continuous application availability in the event that one of the data centers goes
down.
• HP Metrocluster—HP Metrocluster replicates data with array-based mirroring products such as HP
Continuous Access XP, HP Continuous Access EVA, or EMC Symmetrix Remote Data Facility, which
allows the clustered servers to connect only to their local storage devices. Metrocluster provides
automatic site failover and failback for up to 16 HP 9000 servers or HP Integrity servers located up
to 300 kilometers apart.
• HP Continentalclusters—HP Continentalclusters is the most flexible of the HP-UX disaster-tolerant
solutions. Continentalclusters, which permits a pair of Serviceguard clusters to be linked across any
distance, supports local failover within a cluster as well as remote failover between clusters. It
supports both array-based and software-based data replication, including HP Continuous Access
XP, Continuous Access EVA, EMC SRDF, and Oracle Standby Database. The Continentalclusters
solution also provides disaster protection for Oracle RAC environments over unlimited distances.
• HP Mirrordisk/UX, HP Veritas Volume Manager (VxVM)—HP Mirrordisk/UX and HP VxVM enable
data replication to help meet high availability and disaster recovery requirements. HP
Mirrordisk/UX helps prevent data loss by maintaining up to three copies of data on separate disks;
HP VxVM allows up to 32 copies. Both Mirrordisk/UX and VxVM work with the HP-UX kernel to
manage the mirrored disks; hence, they are transparent to the applications and the applications
require no modification.
Disaster-tolerant solutions for Linux
HP offers the following disaster-tolerant solutions for the HP ProLiant and HP Integrity server families,
all of which are tightly integrated with Serviceguard for Linux:
17