Skip to main content

I’ve created a basic workflow that utilizes the API to execute an existing project/plan.  The project had an post process component that writes the invalids to an Excel spreadsheet.

The other steps in the workflow wait for the file to be created and load it into DQIT.  We can write directly to DQIT due to an issue with Databricks/Spark.

 

The workflow works great, but I need to manually execute it.  I’d like to us Admin Center to “Scheduler” to execute the workflow.

How do I do that?

The environment is Azure / Linux.  

Hi, you need to prepare a scheduler, which is an XML file that you have to deploy to the runtime server (same server where you have the workflows). We have a documentation: https://support.ataccama.com/home/docs/aip/latest/development-guides/one-runtime-server-development-guide/workflow-and-scheduler/scheduler There is an example for scheduler executing a workflow, which is exactly what you need. As to where to place the file then, you have to place it into a folder, that is specified in the server config, in the Scheduler component: https://support.ataccama.com/home/docs/aip/latest/development-guides/one-runtime-server-development-guide/server-reference-guide/server-components/scheduler-server-component

There is also an example of schedulers in the Workflow Tutorials in the IDE. You can use it as a template.

Let me know if you need more information.

Anna

 

📌  Links to the legacy documentation portal are no longer available. Read about the changes and how to access our documentation here.


Hi @anna.spakova ,

I have a similar issue at hand having version 14.2 and Hybrid deployment in our case. 

While I nderstand the concept how to create a workflow and server/runtime configuration and also have gone trough the tutorial, something still eludes me. I can create the schedule in a similar fashion to the tutorials, but since it is not a self managed install I understand the .sch file and the .ewf file has to be on the same machine as the Runtime Server is and not to be on the local personal laptop. My question is, which is unfortunately not detailed in the documentation or here in the community is, how I can ‘copy’ / push the components to the runtime server in the first place?

Thansk, Viktor


Hello @ViktorSzucs if you are hybrid, I assume that means that most of the platform is in Ataccama PaaS. In that case, you should have been provided with a Git repository:

The repository works for deploying files to the orchestration (Runtime) server as well as for RDM/MDMPDQIT if you have those too. There are three branches - dev/test/prod - each one for one of your PaaS environments.

There should be a folder for orchestration (Runtime) server (orch/Files), where you can place your plans, workflows, schedulers etc. Also the configuration is available under runtime folder where you can edit pieces of the runtime configuration like sources, path variables and so on. The folder structure shouldn’t be changed (e.g. folder for schedulers) beause it is configured like this in the server configuration.

Once a code is deployed, you need to restart the orchestration service through the Portal for the code to take an affect:

 The Portal should be availale for dev and test environments, production restarts need to be requested through our support team.

If you are not aware of this repository, please contact either your Ataccama contact or our support team to provide more information. I will also check with the documentation team if we have this explained in the official documentation.

Kind regards,

Anna


Hi @anna.spakova 

many thanks for the quick response. I reached out to support to get the Git password already, I also assumed it might be required. Could you please direct me, where the self-service portal is normally located? Currently I only would like to do it on he DEV  https://{customer}.dev.ataccama.online/


Hello @ViktorSzucs , the portal URL should be: https://portal.<customer>.dev.ataccama.online/

Kind regards,

Anna


Hi @anna.spakova 

many thanks for the answers. I have managed to contact support and get the most of the issues sorted out. But now I am stuck again and unfortunately support seems to have a long turnaround time.
 

Maybe you can point me in the right direction. Now by editing the pathvariables.xml I can setup the shortcuts, but here comes my problem.

I have also managed to upload some of the components we have created to the Git repository in dev branch. It seems the pathvariables.xml only takes absolute paths, where I see to struggle to find the correct location for instance to the  dev / customer / orch /runtime 
or any other path within the repository such as below. What would be the default full path to this folder?

Thanks, Viktor


Hello @ViktorSzucs, AFAIK the pathvariable should start with:

/srv/conf/server/orch/

So for the Files/data in the orch folder it would be e.g.

/srv/conf/server/orch/Files/data

This would definitely be better to confirm with the support team but before they answer, you can try this.

Kind regards,

Anna


Thanks for the help. For future reference the following worked out eventually

the {org}/orch/runtime/pathvariables.xml to be setup as so:

<pathVariable name="WORKFLOWS" value="/srv/conf/server/orch/Files/workflows"/>

Additionally the {org}/orch/runtime/workflows.xml on GiT had to be edited with following in order for the runtime server to see the actual files

<source>
<path>./orch/Files/workflows</path>
<id>WF1</id>
</source>

 


Thank you for sharing your solution @ViktorSzucs, much appreciated 🙌🏻


Reply