Managing OpenLMIS’ Configuration Data

(Ben Leibert) #1

Hi everyone,

In order to efficiently load country-specific data into OpenLMIS, Malawi’s implementation team created the openlmis-refdata-seed tool. Because this was prior to the existence OpenLMIS’ stock management service, the tool didn’t include support for data which resides within it.

Because Angola’s implementation of OpenLMIS will use stock management from the get-go, it would likely be useful to add support for stock-related configuration to the tool. This would potentially include support for:

  • StockAdjustmentReasons
  • Valid Sources/Destinations

User management has also continued to evolve since the initial creation of the openlmis-refdata-seed tool. Adding a user now requires the creation of configuration within multiple services, and the refdata-seed tool’s current support for user-creation is therefore insufficient.

For these reasons, I suggest that we start thinking of the utility more generically as the “data” rather than “refdata” seed tool. With that in mind, would the Core team be amenable to:

  • Pull-requests which fix user-creation, even though doing so involves services other than ref-data.

  • Pull-requests which add support for StockAdjustmentReasons and valid sources/destinations within stock-management.

  • Perhaps generalizing the tool’s name.

I imagine the first two of the above changes would be innocuous. I understand that JSI is building a UI for the tool, though, and don’t have a sense of whether tweaking its name would meaningfully effect them.

Thank you,

Ben

(Josh Zamor) #2

Or as the title of your post suggests, rather than just an overly vague “data” tool, we’d like to have a configuration tool.

Yes to all.

The most damaging thing that that team, or others would encounter for your proposed changes, is if we renamed the repository. Lets keep the repo name, and simply/briefly explain the name change somewhere in the README.

Thanks @Ben_Leibert1!

(as an aside this is probably relevant to the #developers topic, though given this is the first content for #implementers there’s no harm in having it here)

(Ben Leibert) #3

Thank you so much, Josh. I renamed the tool as per your suggestion within the readme, and left its repository’s name unchanged.

Do you happen to know whether JSI is already planning to update the tool’s support for user creation? I don’t expect them to add support for new data types, but understand they’ll potentially fix existing ones.

(Josh Zamor) #4

Good question @Ben_Leibert1. I don’t know and atm I don’t see a bug report/issue for that. @Elias_Muluneh what do you think?

(Elias Muluneh) #5

@joshzamor and @Ben_Leibert1, The repository openlmis-refdata-seed was last updated a while back. I anticipate that there could be some APIs in OpenLMIS that have changed since. One of the tasks we are planning to do in relation to the upload reference data work is to go through all data uploads that are currently supported by the refdata-seed tool and verify if it works.

I would not anticipate we would work on supporting a new data upload type but for existing uploads that may have been broken like user creation, we plan to fix.

(Josh Zamor) #6

Thanks @Elias_Muluneh.

@Ben_Leibert1, I think the thing to watch out for here is trying to coordinate Angola’s needs with any potential work @Elias_Muluneh’s team might be taking on. I think Anogla’s timeline is likely much shorter than anyone else. I think the place to start is to add these issues, to the best of your knowledge Ben, into OpenLMIS’ Jira so that we can start planning a release(s) for that component.

(Ben Leibert) #7

Thank you very much, @joshzamor. I sent you a separate note a moment ago regarding how best to share tickets between the Core and Angola.

In addition to the issue with Users mentioned earlier, @Elias_Muluneh, I think you’ll find that the tool’s support for FacilityTypeApprovedProducts needs a bit of work. It’ll initially add them, but doesn’t always remove all of them as intended when run for a second time.

Will you please give me a sense, @Elias_Muluneh, for when you’re planning to begin and finish work on the tool?

(Elias Muluneh) #8

@Ben_Leibert1 I have started working on it. The goal is to complete the upload reference data work by release v3.6. I think there will be significant progress on this in the current sprint.

(Ben Leibert) #9

Thank you very much for the heads-up @Elias_Muluneh. Just in case it’s helpful, the comments within https://openlmis.atlassian.net/browse/AO-78 describe what turned out to be the minor issue regarding FacilityTyepApprovedProducts.

Kind regards,
Ben