Introduction
SSH
A An SSH solution works with a JobScheduler Master that uses SSH to execute commands directly on remote machines as shown in the following diagram:
...
| Feature | SSH | JobScheduler Agent |
|---|---|---|
| Execute jobs with different user IDs | yes | yesCentral job configuration |
| Configure jobs centrally | yes | yes |
| Internal Provide API jobs available, see JITL Jobs by topic, see API Interface | no | yes |
| Real Provide real time log | no | yes |
| Control termination of task terminationtasks, see How to Terminate Tasks | no | yes |
| Distribute job load | no | yes |
| Fail-over for high availability, see Agent Cluster | no | yes |
| Run Runs without installation | yes | noOperating system independent |
| Run on all operating systems | no | yes |
Please note the following:
- Mixed operation of SSH and JobScheduler Agents is possible.
- Further that the addressing of the server on which the job should be run in both procedures is different. Jobs to be executed by a JobScheduler Agent are assigned through the Process Classes and the jobs run under the user ID and the privileges of the JobScheduler Agent.
- SSH jobs have parameters to define the server as well as the authentication.
- SSH jobs require a minimum of 32MB main memory per job execution for a JVM to be loaded. Consider memory consumption for higher number of parallel SSH jobs.
- SSH can be used under Windows with copSSH (http://www.itefix.no/i2/copssh), an openSSH implementation, or similar products. Windows environments provide limited support for SSH, see Limitations when using SSH jobs for Windows.
...