cancel
Showing results forย 
Search instead forย 
Did you mean:ย 
Data Engineering
Join discussions on data engineering best practices, architectures, and optimization strategies within the Databricks Community. Exchange insights and solutions with fellow data engineers.
cancel
Showing results forย 
Search instead forย 
Did you mean:ย 

Anyone using Kinaxis Data Publish (Delta Sharing) from Databricks Marketplace?

vamsi_simbus
Databricks Partner

Hi all,

We're looking at Kinaxis Data Publish on the Databricks Marketplace. It shares Kinaxis Maestro planning data into Databricks through Delta Sharing. We requested access a couple of weeks ago and haven't heard back, so I'd like to hear from anyone who has used it.

Data: What kind of data does Kinaxis send through it?
Direction: Is it a one-way transfer from Kinaxis to Databricks, or can data be sent back into Maestro?
Access: Do you need an existing Kinaxis Maestro licence, and how long did approval take?
Any experience or pointers would help. Thanks!

2 REPLIES 2

Isaac_18
Contributor II

Hi @vamsi_simbus, I haven't used this specific listing, but I can answer the parts that come down to how Marketplace and Delta Sharing work, which should help while you wait.

Direction. Delta Sharing is one-way and read-only. The provider (Kinaxis) shares tables, you mount the share as a read-only catalog in Unity Catalog, and you query it. There's no write path back to the provider through the share. If you want results to go back into Maestro, that has to happen some other way, usually through Maestro's own integration/import options, and that's a question for Kinaxis rather than Databricks.

Access and approval. For listings that say "Request access" (rather than instant access), the provider approves the request, not Databricks. So the delay is on the Kinaxis side. Since the data is your own Maestro planning data, I'd expect them to need an existing Maestro customer relationship to tie the share to the right tenant, but I'd confirm that with your Kinaxis account team. Pinging them directly with the date of your Marketplace request is probably the fastest way to get it moving.

What data comes through. The exact tables are defined by Kinaxis, so the listing page and their docs are the source of truth. Once you're approved, you can check it yourself in a couple of minutes:

SHOW SCHEMAS IN kinaxis_share;            -- whatever you name the catalog
SHOW TABLES IN kinaxis_share.<schema>;
DESCRIBE TABLE EXTENDED kinaxis_share.<schema>.<table>;

One practical tip: treat the shared catalog as your bronze layer. Grant read on it to a small group, and build your own silver/gold tables from it in your catalog, so your pipelines don't break if the provider changes a column.

Hope that helps, and if you do get access, it'd be great to hear back what the datasets look like.

anuj_lathi
Databricks Employee
Databricks Employee

Short version: Kinaxis decides what goes into the share and who gets approved, and from the Databricks side the share is read-only. I can't give you first-hand Kinaxis-specific details, but the platform mechanics answer most of your questions.

Data

What's in the share is defined entirely by Kinaxis. Marketplace datasets are typically published as catalogs of tabular data, and providers can also share volumes, notebooks and models [1]. For Kinaxis, I'd ask them for the data dictionary or schema list for the Data Publish share. Which Maestro planning objects are exposed (and how often they refresh) is a Kinaxis-side configuration, so their documentation or CSM is the source of truth, not the Marketplace listing blurb.

Direction

On the Databricks side this is one-way. A share is a read-only collection of tables and partitions that the provider exposes to recipients [2], so you'll query the data in Unity Catalog but can't write back through the share.

If you want to push results back into Maestro (say, ML-generated forecasts), that needs a separate integration path with Kinaxis, such as their APIs or import mechanisms. Ask Kinaxis what they support there. Nothing about Marketplace or Delta Sharing (now documented under the name OpenSharing) provides a return channel [1][2].

Access and approval time

  • Marketplace listings come in two flavors: some are instant once you accept the terms, others require provider approval and completion through the provider's own process [1]. A request that has sat for two weeks sounds like the second kind, sitting in a manual queue on Kinaxis's side.
  • Whether a Maestro licence is required is Kinaxis's commercial decision. Since this is a feed of Maestro planning data, I'd expect it to be tied to an existing Maestro customer relationship, but confirm that with them.
  • I wouldn't keep waiting on the Marketplace request alone. Go to your Kinaxis account rep or CSM, tell them you've submitted a request from workspace X, and ask them to approve it and enable Data Publish for your tenant.

Practical checklist

  1. Submit the request from the Marketplace in a Unity Catalog-enabled workspace. Requests have to be made from a workspace, even though you can browse the open Marketplace without one [1].
  2. Make sure whoever requested it has the right Marketplace/metastore privileges, and that the request references the workspace/metastore you actually want the data in.
  3. Ask Kinaxis for the expected approval SLA, the data dictionary, the refresh cadence, and any write-back options.
  4. If you're not on a Unity Catalog workspace, shared data can still be read through external tools like Spark, pandas or Power BI [1][2], but Unity Catalog gives you the cleanest governance.

If you've got a Kinaxis contact, escalate through them first. That's usually what moves a stalled approval.

References

[1] What is Databricks Marketplace? | Databricks on AWS โ€” https://docs.databricks.com/aws/en/marketplace
[2] What is OpenSharing? | Databricks on AWS โ€” https://docs.databricks.com/aws/en/opensharing

Anuj Lathi
Solutions Engineer @ Databricks