Implementing Jazz application HA using VMware HA

High availability (HA) at the application tier for the CLM and SSE products can be achieved leveraging VMware technology and tools. VMware technology and tools permit hosting and managing server assets in private and cloud deployments that are deployed on x86 hardware architecture.

Assumptions

• CLM and SSE topology uses a reverse proxy or other front-end hardware or software load balancer (links)
• Topology has each CLM and SSE app on separate servers with distinct names (links)
• Hardware and operating system are x86 compatible

Approach

At a high level, these are the steps most of our VMware customers follow:

• Have ready a standby client virtual machine
  • This may be achieved through cloning the virtual machine from within the VMware tools
• Monitor the active VM and set triggers / alerts in case of failure
  • This may be achieved through monitoring in the VMware toolset, or using other monitoring tools
• Provision (activate) the standby VM
  • This may be achieved automatically with the VMware toolset or manually (using scripts or actual human intervention)
• Configure traffic to new VM
  • This may be achieved automatically with the VMware toolset or manually (using scripts or actual human intervention)

Implementation suggestions

• VMware vSphere ESXi can manage virtual machines on a single hypervisor. In this context, a standby VM needs to be readied. This can be done through cloning the VM so that a separate server is ready but not active. There may be manual steps to activate the prepared VM and to ensure the front-ending load balancer routes traffic to the new VM.
• VMware vSphere vMotion technology permits creating a cluster or group of multiple hypervisors whose resources are managed as a pool. Virtual machines may move transparently from one hypervisor to another without apparent end-user involvement or awareness. This technology compensates for variability in a hypervisor cluster due to changes in virtual machine load or unexpected hardware issues at the hypervisor level.
• VMware High Availability (HA) technology automatically restarts the virtual machines elsewhere in the hypervisor cluster.

Jazz application recovery

The Jazz applications can be monitored through external tools which check uptime of the actual application or the virtual machine hosting the jazz application.

Scripts can be written and can be run from a system outside of the virtual machine hosting the jazz application.

WAS recovery

WAS can be monitored through external tools to check queues and threads. Scripts can be written and can be run from a system external to the virtual machine running WAS to determine health.

Single virtual server recovery

VMware vSphere console can monitor the virtual machine to determine health and alert or respond proactively.

Other considerations

If there are additional specific failure scenario, please add a new scenario to the parent page and new section within this page

Related topics: Approaches to implementing high availability and disaster recovery for Rational Jazz environments

External links:

• VMware
• VMware vSphere ESXi
• VMware vSphere vMotion
• VMware vCenter Site Recover Manager

Additional contributors: None