While primary control of Qube Remote Control is the Supervisor, the Worker is responsible for making monitoring the remote host and executing jobs dispatched by the Supervisor. Like the Supervisor, once the Worker is configured, there is little need to return to its configuration.
Workers can be configured locally on the each system or centrally from the Supervisor.
Local Configuration
One can configure all Worker parameters locally. However, with You can do this by running the WranglerView UI on the Worker, and going to the Administration menu item "Configure Local Host". Or you can edit the local qb.conf file by hand. The file's location depends on the OS of the Worker:
- Linux & OS X: /etc
- Windows XP/2003: C:\Windows
- Windows Vista/2008: C:\ProgramData\pfx\qube
After making the changes, the WranglerView UI should restart the Worker process for you. If you doing this manually, you must restart it yourself, or run this command:
$ qbadmin worker --reread
Central Configuration
With larger farms, the Worker configuration process can become cumbersome, so Qube! is designed to allow for a minimal configuration local to the host, in favor of a centralized configuration on the Supervisor. The This uses an additional file, qbwrk.conf file on the Supervisor is configuration file is the most efficient place to all your Worker configurations. Since this file is located only on the Supervisor, it allows an administrator to define settings for many Workers in central location.
The qbwrk.conf file is located as shown above, but on the Supervisor. It can be modified by running WranglerView on the Supervisor (as a user with write privileges to the file location) or manually by editing the file directly. If you modify it manually, you must run this command afterwards:
$ qbadmin worker --reconfigure
Autodiscovery
A Supervisor autodiscovery process is used by Qube! as a means to quickly identify the Supervisor without any local configuration necessary. If the Supervisor isn't explicitly set in the Worker's configuration file, it will use autodiscovery to find the Supervisor host name and IP address. However that Supervisor and Worker must be on the same network subnet. Once it has that information it can then gather configuration from the Supervisor and continue the startup configuration process.
In general, it is preferable to explicitly name the Supervisor however, since this speeds job startup time and reduces network traffic slightly.
Other Configuration Topics
| Children Display |
|---|