Production sync done with a bug in DEV Model

Hi,

When we make a Production move and down the line 2 weeks, we get to know the production moved with an error, and there is lot of data added by the client in this time, what would be the best possible ways to fix the issue?

1. Move the production back to the last revision, fix the Dev and do a sync again? but then there will be lot of data loss.

2. Is it recommended to move your production model back in history through restore ID? Or it should always be to the last revision tag. Is there any disadvantage to restore it to some restore ID, does the Dev and Prod sync gets break doing this?

Thanks.

Accepted answers

All comments

Misbah

@ShivankurSharma

With both these approaches either you will lose the data or you will break the sync. I was thinking of creating new Dev Model from the current Prod Model. Not sure what kind of error are you getting. Can you please detail it out

ShivankurSharma

Thanks @Misbah

Actually a line item was initially a User Input which was pushed as a formula in sync, which triggered the data loss for the line item. To fix that we pushed back the production model to some restore ID (there was a data loss doing this), and then the production and development model ALM got break and then to fix that we had to create a fresh Dev model from the Production and redo the development.

Yes, the data loss was there. But is it if we take the production model back in history through Restore ID instead of Revert to Last revision the ALM breaks?

Misbah

@ShivankurSharma

Yes if you go beyond your last revision tag you will lose the sync. I think whatever you have is inline with what should be done - unless any other champs have different thoughts

ShivankurSharma

yes. Thanks.

kavinkumar

@ShivankurSharma,

Let me put my thoughts and check if it works for you.

This seems to be a critical issue and please be aware at any point before doing the Production sync.

Before you start, please do a COPY and ARCHIVE of the existing Production model to be on the safer side.

Step 1: Take your PROD model out of Deployed mode and go back to the point before the last Revision Tag as you mentioned. (Note: DO NOT MAKE ANY CHANGES)

Step 2: Export the data from the line item (Mentioned by you in the above thread) and keep it in CSV or TXT.

Step 3: Get the model back to its original point and put it in the deployed mode (Make sure the Sync doesn't break)

Step 4: Remove the formula from the line item and Sync it with PROD.

Step 5: Now create a one-time activity action to import the line item data into the module.

I hope this will prevent you from either breaking the Sync or Data Loss. Let me know if you have any concerns with this approach.

ShivankurSharma

Thanks Kavin, I believe this could be a better approach, until and unless the Sync don't break.

Have you ever tried this approach where you change the Production to standard and revert the model beyond the last revision and then reverted back the model to current state and move the model to Production and still the sync remains intact?

kavinkumar

HI @ShivankurSharma,

I tried replicating the scenario in my test model and seems to be working fine. Ideally, it should work!

Does anyone disagree?

ShivankurSharma

Thanks Kavin,

Community Members can you confirm is you foresee any issue in this approach or if we can take it as an accepted solution to fix the Production Model.

Misbah

@kavinkumar

I don't disagree but it is highly highly risky. Sometimes taking the model out of deployment mode breaks the sync without even making any change. Also it is very important to remember which RT has one to fall back to before putting model back into deployed model.

Misbah

@ShivankurSharma

Since the sync is already broken in your case new Dev model is required after all. Which brings me to the point how Important it is to use Copy & Archive option to Archive your models on a daily basis as part of your Disaster Recovery Process.

Since the intention is to get back the input values from an end user you would have unarchived that model and take out those numbers and load those into you current production model.

jasonblinn

Based on your description I would recommend this approach: https://help.anaplan.com/anapedia/Content/ALM/Compare_and_Synchronize/Fixing_Production_Issues.html?_ga=2.153328955.2035036762.1586793311-1876461626.1579557669

Essentially you roll back DEV to a prior state before the bug was implemented. Then create a revision tag and Sync to PROD. Then Roll DEV Forward to before you just rolled it back, Then you can keep developing in DEV and you have returned PROD to a prior state that is prior to your bug.

Production will not allow you to roll back past your previous Sync without changing the mode and destroying the ALM Compatibility between models.

Jason

ShivankurSharma

It's true. It will be highly risky.

ShivankurSharma

Thanks Jason.

ShivankurSharma

Yes we moved with the same approach. Did a hot fix and added the data again.
Thanks for the help.

Quick Links

Can you help?

Unanswered questions

Importing Ragged Hierarchies
Hello - I have a 2 level ragged hierarchy (yes, it's a chart of accounts), that for reporting purposes is used in 4 different models (yes, it's a P&L). What is the best way set this to allow one to be the "source of truth" and then import between the four models without creating actions that include ignores? It seems like…
Where can I find export file after running export process using machine task in a workflow?
I am testing export actions using workflow but its not generating any file when I am running the workflow, but the same process is generating export file when ran from the Actions→Process section. Is their any issue with the workflow functionality? Also, If I want to generate a file after completion of workflow and pull…
Automating Flat File Uploads to Anaplan Data Orchestrator (ADO) – Best Practices and Methods
Hello Anaplan Community, I am seeking guidance on the most effective ways to automatically upload flat files to Anaplan Data Orchestrator (ADO). Currently, our process involves loading flat files to an Anaplan Datahub using Anaplan APIs and custom Python scripting. With Anaplan Data Orchestrator's capabilities for data…