Server Talk

Hi Support Team,

whats the best way to let two 24x7 Scheduler(Master) "talk" together.
JobRemoteRun does not have a result String. So how do i get a feedback?
Is this possible to start a job on the other side with a result as return - like a funktion. Or is it only a oneway ticket.

Best regards
Dominik

Hi Dominic,

What kind of a value do you need (some calculated value, or just process output, or something else) and how is it going to be used?

Different solutions are available for different needs. the most common is to have the agent execute a job and have the job save data to some file(s). the use the built-in secure file transfer functions to transfer files from the agent to the scheduler. See "File.transfer" for more details.

Maybe i dont understand the way JobRemoteRun works.

I start it on Server A --> JobRemoteRun ( 1, ServerB)

Will the JobId 1 of A start on B, or will the JobId 1 of B start on B?

At the moment I get the error: Unable to run job 1 remotly. User credentials not specifierd for that job.

	Quote:
	JobId 1 of A start on B

Job definition created on server A in run time will be deployed to server B and executed there.

if you run 24x7 v5, please do not use old legacy JAL JobRemote... commands. you should convert the script to JavaScript (it will use Scheduler.runRemoteJob) or statically assign Job 1 to run remotely and then use JobRun. In the latest case just set job's host property to a profile configured for the 24x7 agent running on server B

	SysOp wrote:
	Job definition created on server A in run time will be deployed to server B and executed there.

Ah, ok thats explains much.
But what about the queues? Where is the A-Job defined in B environment? (Listed in, Prio etc)

	Quote:
	But what about the queues? Where is the A-Job defined in B environment? (Listed in, Prio etc)

The A-job is not defined in B environment. The B-agent is like an remote arm of the A-scheduler or any other scheduler. It executes jobs and commands sent by schedulers. An instance of A-job is created somewhere in B-agent memory during job execution, there is no queue or any other storage associated with it.

By the way, queues are not related to any storage, queues are job execution engines allowing the scheduler to run multiple parallel job streams with user specified sequencing, queuing, and priorities. Agents don't have queues, they are remote automation agents that do things they are told to do by the schedulers exactly when they are told to do it (no queuing).