This section describes how to manage both planned and unplanned outages of vFabric tc Server.
In a planned outage, you schedule a time when tc Runtime instances will be briefly unavailable so that you can perform maintenance on the instance or deployed applications, create cold backups, and so on. The procedure describes how to stop all tc Runtime instances.
If you are using a Web Server as a load-balancer or proxy in front of one or more tc Runtime instances, drain all currently opened sessions between the Web Server and the tc Runtime instances.
For example, if you are using vFabric Web Server, you can simply
stop the instance using the
httpdctl command, as shown in
the following Unix sample:
prompt# cd /opt/vmware/vfabric-web-server/myserver/bin prompt# ./httpctl stop
In the preceding example, the vFabric Web Server instance is
located in the
stop command forcibly ends all sessions. To
specify that you want the Web Server instance to wait until all
sessions end gracefully, use the
prompt# ./httpdctl gracefulstop
On the computer on which the tc Runtime instances are installed, stop all instances. For example, on Unix:
prompt$ cd /opt/vmware/vfabric-tc-server-standard prompt$ ./tcruntime-ctl.sh -n /var/opt/vmware/vfabric-tc-server-standard myserver stop
In the preceding example, vFabric tc Server is installed in
/opt/vmware/vfabric-tc-server-standard, the name of the
myserver, and the instance directory is
See Starting and Stopping tc Runtime Instances for additional details, such as Windows instructions.
When you stop tc Runtime instances, the Web applications that are deployed to the instances are not available to users. You can now safely perform maintenance on the instance, such as update its configuration and create a cold backup.