High Availability
On-demand failover makes applications highly available
The Red Hat® Enterprise Linux® High Availability Add-On
provides on-demand failover to make applications highly available. It delivers
continuous availability of services by eliminating single points of failure.
Failover
Configure the High Availability Add-On for most applications that use
customizable agents and for virtual guests. By offering failover services
between nodes within a cluster, this Add-On supports high availability for up to
16 nodes. (Currently this capability is limited to a single LAN or datacenter
located within one physical site.)
The High Availability Add-On also enables failover for off-the-shelf
applications, like Apache, MySQL, and PostgreSQL. Any of which can be coupled
with resources like IP-address and single-node file systems to form highly
available services. You can also easily extend the Add-On to any user-specified
application that's controlled by an init script per UNIX System V (SysV)
standards.
How it works
When using the High Availability Add-On, a highly available
service can perform a failover from one node to another with no apparent
interruption to cluster clients.
The Add-On also ensures absolute data integrity when one cluster node takes over
control of a service from another cluster node. It achieves this by promptly
evicting nodes from the cluster that are deemed to be faulty using a method
called fencing, which prevents data corruption. The High Availability Add-On
supports several types of fencing, including both power-based and storage area
network (SAN)-based fencing.
[Features]
The High Availability Add-On for Red Hat® Enterprise Linux®
provides continuous availability of services by eliminating single points of
failure. This article provides the key features and benefits of this Add-On.
Features
|
Benefits
|
Clustering |
Red Hat's High Availability Add-On
enables applications to be highly available by reducing downtime and
ensuring that there is no single point of failure in a cluster. It also
isolates unresponsive applications and nodes so they can't corrupt
critical enterprise data |
Conga
|
The Conga application of Red Hat
Enterprise Linux provides centralized configuration and management for
the High Availability Add-On |
Corosync |
Corosync is a cluster executive
within the High Availability Add-On that implements the Totem Single
Ring Ordering and Membership Protocol, delivering an extremely mature,
secure, high-performing, and lightweight high-availability solution
|
Integrated virtualization |
Virtualization is pervasive
throughout today's enterprise datacenters. Not only is Red Hat
Enterprise Linux designed to be a superior guest on any of the major
hypervisors, but it can also be a virtualization host. Virtualization is
integrated directly into the Red Hat Enterprise Linux kernel using
kernel-based virtual machine (KVM) technology. As part of the kernel,
your administrators get the complete breadth of Red Hat Enterprise Linux
system management and security tools and certifications |
Fencing & unfencing |
Fencing is removing access to
resources from a cluster node that has lost contact with the cluster,
thereby protecting resources such as shared storage from uncoordinated
modification. Red Hat has made extensive improvements in the SCSI-3 PR
reservations-based fencing. By enabling manual specification of keys and
devices for registration and reservation, cluster administrators can
bypass clvm and improve configuration and system flexibility. After
fencing, the unconnected cluster node would ordinarily need to be
rebooted to safely rejoin the cluster. However, unfencing allows a node
to re-enable access when starting up without administrative
intervention. |
Improved cluster
configuration system |
The cluster configuration system now supports
load options other than XML, including the Lightweight Directory Access
Protocol (LDAP). Configuration reload is validated and easily
synchronized across the cluster for better usability and manageability.
|
Virtualization
integration |
You can now run virtualized KVM guests as managed
services |
Rich
graphical user interface (GUI)-based cluster management and
administration |
In Red Hat Enterprise Linux 6, the Web interface to luci has been
redesigned and runs on TurboGears2 |
Unified logging and
debugging |
System administrators can now enable, capture, and
read cluster system logs via a single cluster configuration command
|