SAC Content Network - promote Data Import Jobs to Production

[sap_cohort]

Hi, Is there a way to export/import "Data Import Jobs" from development to production?
I transported Connections and Models and none of the related Import Data Jobs came over.
Can't believe I would have to recreate all these jobs I have.
Is this possible?
Also, can't find any info on this on the internets. Thanks!

[sap_cohort]

Thanks for the response. This is all very disconcerting.
I am looking to export/export just the import jobs on a specific model for now. Nothing for dimensions.
When I bring up the Deployment or Content Network without selecting anything I see under Wrangler a total of 27 items (unselected) with no text just id numbers.

I have at least 60+ import jobs so I'm trying to figure out how 27 "Wrangler" items can equate to all the import jobs I have?
When I select a specific model none of the "Wrangler" items get selected so it looks like I would have to manually select them.

I guess I don't understand why something as simple as wanting to promote an Import Job with a name can't be identified and imported/exported. And there is no information on this topic other than the note I found listed below which does not add much value. I can't possibly be the only one having to do this so I'm not sure why I'm unable to find any info on this. Unless everyone else is just building their import jobs in the production environment. I have so many import jobs with mapping rules it would be really challenging to do so.

Would you be able to add a little more context to importing/exporting Import jobs and the wrangler topic equating to import jobs?
Thanks a lot for your help!

2800832 - How-to: About Wrangle contents shown in Export list of Deployment in SAP Analytics Cloud (SAC)
https://apps.support.sap.com/sap/support/knowledge/en/2800832

[mfoeken]

Hi Kenneth,

It's possible to preserve the import jobs when importing a model into a production SAC tenant. However it's not completly clear what the prerequisites are.

Are you the owner of the connection in both source and destination tenant? Is the connection imported on the destination tenant or was it originally created seperatly on both tenants?Try to import a model without the connection.

Kind regards,

Martijn van Foeken | Interdobs

[DebjitSingha]

Hello kenneth.murray2,

How are you.

Query jobs fall under "wrangler". If you pick wrangler along with dimensions then you will see query jobs gets imported / exported along with dimensions. When you select models (from File), system selects wrangler by default.

Content Network Screen:

Content network option shows that wrangler is selected. Once you click on Wrangler, system show the associated ID.

Concept is same if you use Deployment (Export/ Import) option as well.

Deployment Screen:

Hope this helps. Let us know if you have any query around this topic.

Thanks,

Debjit

[DebjitSingha]

Hello kenneth.murray2,
Please find my comments below:

Thanks for the response. This is all very disconcerting.
I am looking to export/export just the import jobs on a specific model for now. Nothing for dimensions. You can do this by first selecting models, system will select corresponding (only required ones) wrangling components > Uncheck everything accept wrangler
When I bring up the Deployment or Content Network without selecting anything I see under Wrangler a total of 27 items (unselected) with no text just id numbers. Wrangling is a background component, we can not even view this component out side of import export process. Hence only ID's no text/ description.

I have at least 60+ import jobs so I'm trying to figure out how 27 "Wrangler" items can equate to all the import jobs I have? Wrangler are specific to models. One wrangler per model. Say you have 10 jobs with Model 1, 1 associated wrangler will consider all 10 jobs in it.
When I select a specific model none of the "Wrangler" items get selected so it looks like I would have to manually select them. System picks depended object from top of the hierarchy. So in this case if you select models (under Files>Public>......) then system will select dependent wrangler and other objects along with it. No need to do manual selection. You can even export all components to tgz file and while import just select wrangler and uncheck everything else.

I guess I don't understand why something as simple as wanting to promote an Import Job with a name can't be identified and imported/exported. And there is no information on this topic other than the note I found listed below which does not add much value. I can't possibly be the only one having to do this so I'm not sure why I'm unable to find any info on this. Unless everyone else is just building their import jobs in the production environment. I have so many import jobs with mapping rules it would be really challenging to do so. Just follow the steps from Model and then backwards. Export everything (associated objects) if, your are not sure. Then while import just select wrangler.

Would you be able to add a little more context to importing/exporting Import jobs and the wrangler topic equating to import jobs?
Thanks a lot for your help!
2800832 - How-to: About Wrangle contents shown in Export list of Deployment in SAP Analytics Cloud (SAC)
https://apps.support.sap.com/sap/support/knowledge/en/2800832

Let us know if you have further question....

Thanks,
Debjit

[sap_cohort]

Hi, Thanks for all the info! Greatly appreciated!
Makes sense now that 1 wrangler would be associated with a model. Wishing this part was made a little more clear for everyone. Although, when I select my model for export none of the wranglers are selected....
Any thoughts on this and how I might be able to identify the correct wrangler to export?

Thanks again...

[DebjitSingha]

Hello kenneth.murray2,

Try below steps: (start with one model, once you are comfortable add more in one package)

Your Source SAC Tenant > SAC Menu > Deployment > Export > Click on + sign > Do not select anything yet - expand File at the bottom till you see one of your model > Just select the check box next to your model > System should pick up wrangler on right panel > Export (do not uncheck anything yet) > Save tgz file localy

Go to your target SAC tenant > SAC Menu > Deployment > Import > Click on upload icon (next to + sign) > Select source file from your local folder > Upload > Now uncheck everything except wrangler > Click Import on top > check Update existing object > Ok
Go back to your model

Let me know how it goes.

Thanks,
Debjit

[sap_cohort]

Thanks again for your help with this. This seems to be buggy.
I did exactly as you mention and also tried in Content Network. Selected the model alone and no Wrangler items are selected at all out of the 27 that are in the system! Wondering if this is because the Model has already been sent via Content Network without the Wrangler.
This is my largest model with lots of Import Jobs so I would imagine it would have an associated Wrangler.

Would you have any idea where to go from here to get the wrangler?
I'm hoping to get this right so I don't have to recreate all the jobs in Prod!
Wondering if there is a way to see the Model metadata to identify the associated Wrangler?

If we do get this to work and the Wrangler does get sent to the Prod System what would be the impact of resending the Wrangler?
Would the 2nd Wrangler overwrite any changes made to Prod Import Jobs with the old QA Job Import Definition?

[sap_cohort]

Also the OSS Note doesn't mention anything about Wrangler being related to Import Jobs.
2800832 - How-to: About Wrangle contents shown in Export list of Deployment in SAP Analytics Cloud (SAC)

[DebjitSingha]

When ever you create a import job a wrangler object is created. You can view it in activity log.

If your system is not auto selecting dependent objects then its a bug and you should open a case with sap support. This is odd behavior. Regardless if you are LCM process involves content network or deployment objects should auto detect dependents.

Assuming you may not have done any direct changes in production, only risk of reimporting of wrangler is resetting of delta pointer ( if your import jobs are based on delta updates). This can happen if you set “overwrite” during import. Say you created new jobs in production only and now import wrangler without overwrite set- then after import you will see both jobs (production set and dev set). In short if you do not update the dev jobs will get added below prod jobs.

If you didn't change model directly in production, then you can move wrangler. In my opinion fixing job delta pointers are easy task than creating transformation all over again. If you know what I mean

Thanks,

Debjit

[sap_cohort]

Hi Thanks for the follow-up. Not always hitting up the Community so apologies for the delay.

My last comment is pretty much where I am:

I just noticed that when I select a Wrangler object it does seem to auto-select the dependencies. So I went through each of the 27 Wrangler objects I have and not one selected the Model I am interested in.. I'll try again to see i can notice any pattern.

Thanks for sticking with us on this!

[DebjitSingha]

Hello kenneth.murray2,

What happens if you try to capture models in your target system?

Same behavior or wrangler is selected? If wrangler is selected, then is the ID same as in your source system?

If you see issue across the board (unlike though), then best approach is to create an incident with SAP Support. SAP may take sometime to identify and fix this.

Meanwhile, as mentioned in my last reply, you can overwrite your target wrangler in certain scenarios (explained in last reply) and still get this through.

Here is something you can try, with some assurance that if something goes wrong you have production backup to go to.

Take a backup of the models from target system : Use deployment option and collect all required objects into a local tgz file. This will act as backup. In case you target system is showing same behavior and not resolving dependencies automatically then, try to capture objects in multiple tgz files. More granularity of object better. Manually document objects captured vs request number (tgz file name).

Now move data wrangler(s) from source system to target system and validate (preferably one by one). If something goes wrong, restore production version from local backup (in overwrite mode)

let me know if you have any question. We invested lot of time and effort on this issue, it would be great if we see this through.

Thanks,

Debjit

[DebjitSingha]

Hello kenneth.murray2,

Didnt herd back from your. Wanted to check, did you tried the steps mentioned in last post?

Thanks,

Debjit

[andreasja_schneider]

Similar scenario for us:

WRANGLER objects did not do the job during transport for our acquired SAC Models.

It seems the DATA Job schedules (for acquired SAC Models under tab data management) are linked to the connection, and not under WRANGLER.

[DebjitSingha]

Hello Andreas,

"Connection" do not consider scheduled for individual models (import). "Connection" only consider the linkage between source and target. While importing objects wrangler consider queries and not exactly the schedule (from source system). Once query is imported (part of wrangler) one need to schedule them in target explicitly. At lease this is how I have seen SAC behave when it comes to HANA or BW based import connections.

Thanks,

Debjit

[JefB]

The import jobs to a SAC new model do seem to get transported as part of the model object, so not as part of connection nor wrangler objects.

But you just are required to run the job once in the target system by opening it or by scheduling it, before it becomes active for ad-hoc execution (refresh icon). Hope this helps...

[DebjitSingha]

At the time of question posting (around March) "new model" was not available 🙂 Question was answered, keeping clasical model in mind.
Having said that, yes you are correct new model changes how jobs and queries are captured.

[kanoun68]

We are facing the same issue and also not able to map the wrangler objects to model we exported/import toward production tenant. I am quite curious were you able to solve this issue? Please let us know what you have done!

[sap_cohort]

I just noticed that when I select a Wrangler object it does seem to auto-select the dependencies. So I went through each of the 27 Wrangler objects I have and not one selected the Model I am interested in. The saga continues...