Best practice question

We are upgrading from v3.x to 5.x and the one big difference between the versions is that remote control is no longer available on v5. There are a few people working on the scheduler at our firm and we could all be working on it at the same time. We each could be debugging a job and if I am not mistaken, web console does not have the debugging feature. It seems that the installation mode of "master scheduler with local GUI" fits our needs the best but only one person could be using it at any one time. What's the best practice to work around this situation? We are running in distributed mode with one main scheduler and multiple agents. Should we each have a local install of the scheduler and debug code from there? Any guidance is appreciated. Thanks.

You are correct. The web console is now the main tool providing multi-user access to the scheduler and job management. It doesn't support remote script debugging (yet).

Another toll providing multi-user access is the command line based console.

To debug scripts in graphical mode, one can start the scheduler with /NOTIMER option. This starts a standalone scheduler instance with the timer disabled and all time and event-based processing suppressed.

Yes, I am aware that the scheduler can start with the timer disabled. Can I have multiple users logged on to the scheduler server and start up the scheduler that way without impacting the main scheduler engine? Thanks.

It can be safe only if each user starts a separate copy fro ma separate directory. One way to workaround this is to develop a batch file to copy scheduler files, alter path referenced in preferences.xml and then start user copy of 24x7 in NOTIMERr mode.

Here is a untested example demonstrating the described idea

1. copy contents of c:\24x7 to a user specific directory

	Code:
	@echo OFF echo Please wait C: xcopy c:\24x7 "%APP_DATA%" /E /I /H /R /Y

2. update reference to the job database file with a user-specific reference

	Code:
	CD "%APP_DATA%\24x7" type preferences.xml \| powershell -Command "$input \| ForEach-Object { $p = '%APP_DATA%\24x7' - replace '\\', '/' ; $_ -replace 'c:/24x7', $p }"

3. start in notimer mode

	Code:
	master.bat /NOTIMER

Sorry, I am going to bring this question back to life because we are upgrading to the latest v5 very soon. If we are using "Master Scheduler mode with the local GUI attached" and the scheduler is running as a service, we should still be able to bring up the scheduler GUI on the server, right? I did that and the scheduler started in the no timer mode, which is what I expected. However, I was not able to add a job via the web console when both the GUI and the scheduler service are started. Is that by design? Thanks.

You can bring up the scheduler in GUI mode, just keep in mind there is no separate GUI, it's not a GUI console, it's a full application running in GUI mode. If you do it while the service is running, most likely nothing good would come out of it. So, either use the service, or the GUI. GUI mode is for development, service is for production. If you run both and get web console connect to your half disabled GUI, instead of the service, you won''t be able to create jobs.

The right solution for a production environment, is to use the web console with the service. Use GUI only for special situations only. if you need to troubleshoot or setup something in a graphical mode.