Articles in this section

Distribution job

Overview

Distribution job is designed to distribute data from a server to selected agents and is commonly used for delivering software updates or important documents to remote workstations. Android devices can participate in Distribution job starting with version of Management Console 2.5.0.

How to configure Distribution job

1) Create a new job
Go to JOBS tab -> Configure jobs -> Create new job, pick Distribution, click Next.

2) Give the job a name and description
This is optional and defaults can be used. Check option to use SHA2 hashing if preferred. Note that with SHA2 Agents running Connect version 2.0 and older will not be able to participate in this job. Click next.

3) Pick the source agent
The agent that will be uploading data to others. Only one agent can be picked. Select the path from where this agent will upload files. This directory must already exist and files shall be there before you start the job.

4) Select group of destination agents
If you don't have any, you can create it here and add agents to it. Also, you can separately select some agents for that job, provided the agents do not produce Conflicts. Note, these ones won't be automatically added to any group!. Select the path for these to store files from source agent.
See here for more details about creating and using groups.

5) Specify Triggers
Distribution job supports post commands which make it possible to run a command once transfer is complete. Triggers indicate the moment when the script will be executed. This step can be skipped.

  • Before file-indexing begins
    Right after job is created, the agent will start indexing files in the specified directory. A script, triggered at this moment can "cook the files before serving", for example, re-arrange the them, add/remove new and do things alike, so that the folder is indexed and distributed in a proper manner the way you need.
  • After an agent completes downloading
    The script will run on each destination agent after it finishes download. Other agents may be still downloading the files, thus it's recommended not to remove or update the distributed files with this trigger. Otherwise the file will be re-downloaded from source again. It's possible to schedule the script for the agent's local time.
  • After all agents complete downloading
    As opposed to the trigger above, in this case script will run only after all destination agents finish downloading all the files. This script will also be executed on the source agent.

6) Job scheduler
defines when the job will be launched

  • Run at - at the preferred date/time (local agents' time);
  • Run manually - job won't start until manually launched with "Start" button;
  • Repeat hourly/daily/weekly - job will run at a specified frequency. It's possible to select the when the first run shall be started (a starting point for schedules run) and when schedule expires.

By default, if time to start next run has come but the previous run hasn't finished yet, active run will be aborted and new one started. There's a checkbox to change this behavior and not launch the new run until active finishes:

Run At: new job run should be started in case the previous one completed no later than ran_at_time + 5 mins. Examples:

  1. job run was launched manually and finished at 11:00. scheduler is "run at 10:56". A new job run starts during a minute.
  2. job run was launched manually and finished at 11:00. scheduler is "run at 10:54". No new run.

Run every minute/hour: start a new job run asap in case job run was completed later than expected schedule time. otherwise default behavior. Examples:

  1. job run started at 11:00 and finished at 11:08, scheduler is "run every 5 minutes". A new job run starts during a minute.
  2. job run started at 11:00 and finished at 11:03, the same scheduler. Wait till 11:05.

Run daily/weekly at specific time: the same principles as in Run At. Examples:

  1. job run started 01.01 at 11:00, finished 02.01 at 12:00. scheduler is "run every day at 11:30". New run is expected 03.01 at 11:30
  2. job run started on Monday at 11:00, finished on Wednesday at 12:03. scheduler is "run on Mon at 11:00, Wed at 12:00". A new run starts during a minute.

7) Settings
Select Job's priority and Job's profile.

8) Notifications
Setup and e-mail notification if necessary. 

Done! Depending on the job scheduler, the job will be launched or shall be started manually. Watch its execution on Job Runs page. 

Related articles