Alteryx Connect Ideas

Hello Alteryx Team,

Would you consider creating a metadata loader for SharePoint? A lot of companies that we are working with have a lot of data stored in SharePoint and they would like to catalog their metadata.

I am aware that it is possible to use the Files Loader when the SharePoint drive is mapped as a mapped drive in Windows Explorer, but the mapping and subsequent scheduling is very problematic, also because of security policies in large enterprises.

Having a SharePoint dedicated loader would be a great benefit.

Thank you very much for considering this idea.

Regards,

Jan Laznicka

Our users should primary look for certified objects. How ever the advanced search does not have the possibility to search for certified objects.

Certified objects should appear in the top of any search.

Today it has no effect on the search if a object is certified. Certified objects should be boosted in the search.

In the administration console, Connect configuration, search boost it should be possible to ad weight to certified objects

We're working through an implementation of Connect - and it appears that every time Connect scans the Alteryx or Tableau environment, it does a full reload of all canvasses / workbooks.

However - we have several thousand tableau dashboards & alteryx canvasses - so this creates a significant delay on running the loaders, and we'd want to run these loaders every few hours so that Connect has up-to-date information (or at most 24 hours out of date).    Running a full export and scan is causing this load to take a very long time to run.

Can we change the default behaviour for all the loaders to use a delta-load rather than doing a full scan - i.e. only pull out assets that have changed since the last load?

 cc: @nganesha @Kosi

We are moving our measure specifications that are in a PDF format to the Connect glossary.  We would like to keep the formatting that exists in the PDF when copying the spec.  However, the format doesn't always copy over.  It would be nice to have the 'Format Painter' option similar to what is in MS Word where you can copy the format from one line to another line.

Hi Alteryx,

currently, when running loaders from the Administration Console - Connections pane, there is no information about the status of the load. Would it be possible to include information such as success/fail message, run length, completion time or next scheduled run time?

This would much simplify monitoring of the scheduled loads and managing the schedule.

Thank you very much.

Regards,

Jan

Hi Alteryx,

we have noticed that when drilling up/down in the "all mode" in Nexus, it is not possible to see all subsequent relations for assets related to the object, where Nexus was opened.

For example, when opening the Nexus view on a business term:

We are unable to drill into the outgoing relationships for the highlighted Tableau dashboard, the only relation we see is its parent.

Would it be possible to show all relations when drilling up/down in the Nexus view?

Thank you very much.

Jan

Hi Alteryx team,

Currently, there is a possibility to 'bulk certify' assets from the front-end using the Bulk operations functionality. Would it be possible to add also the 'bulk ownership' assignment feature?

We are aware that it is possible to assign owners using the Excel bulk upload, however, for a lot of users it is much more convenient and straightforward to do it directly from the front-end application.

This would really simplify the governance of assets in a large source systems like SAP HANAs, where certain users own large amount assets. It would enable data stewards to filter the respective assets using the built-in search functionality and assign the owners for the filtered subsets.

Thank you very much for considering this idea.

Regards,

Jan Laznicka

For assets identified in Connect - we are currently not able to identify an asset as being a UAT version of a Prod asset.   Instead, these are listed as 2 completely different assets with no relationship between them.

This will be confusing for consuming users since they won't know which to use; and wasteful for the people capturing data about this asset since they will need to capture the info several times.

Request: Please create the ability to tag a given Alteryx / Tableau / Database server as "UAT / Dev / Prod", which will automatically tag the asset with this type.   Then allow asset owners to relate the prod and Dev version together as the same thing, but in different states.

Thus, when you search you will find 1 asset with several different states rather than 3 assets.

CC: @DavidM @Arianna_Fuller

Hi Alteryx,

We have noticed that the sequence of objects in the Nexus diagram is not unified for different modes. For example, for user defined references the sequence in the „Joins“ mode is different from the lineage displayed in „All“ mode. The same stands for lineage, the arrows are often pointing in the opposite direction than in the “All” mode.

Would it be possible to unify the sequences of objects in the Nexus view?

I understand that the “All” mode just lists all relations without the actual sequence, but this is quite confusing for the users and it may be more intuitive if the arrows always pointed the same way.

Alternatively, would it make sense to have the arrows on both sides for user defined references?

Thank you.
Jan

Connect has the ability to visually see 1 level of dependancy (in the Nexus view).

For an asset owner - it is very important to be able to see ALL upstream / downstream dependancies to be able to understand impact.   Key here is answering the question "who will be impacted if I change XX"

This should include asset; owner; and depth - preferrably in a tree format.

CC: @DavidM @Arianna_Fuller

Hello Alteryx Team,

We would like to be scheduling the loaders from the administration console, Connections tab. However, when scheduling from there, the loaders are being sent to the Alteryx Gallery under low priority and therefore can easily end up in a queue. As we are scheduling a number of loaders that run for a significant amount of time, it is essential to have the schedules set on precise times and have a certainty that they will not end up in a queue. However, the low priority in the Gallery prevents us from doing that.

It would be great if there was a priority switch directly in the Connect's Connections section, where it would be possible to select how prioritized should the loaders be in the Alteryx Gallery. Also a possibility to select a specific worker would be greatly appreciated.

Thank you very much for considering this idea.

Regards,

Jan Laznicka

Hi Alteryx team,

would it be possible to include a project filter in the Tableau loader? Currently, the loader brings in all the projects from the selected sites. However, we have come across a situation when each department in a company has a couple of projects that they do not want to load.

The idea is that it would be possible to list the names of the projects that won't be loaded. It could be done in a way that all the projects containing the listed words won't be loaded. This would give us the possibility to, for example, not load the projects containing 'Ad hoc' in their names.

We are aware that we might keep these projects 'invisible' using permissions, however, for operational purposes we believe that the project filtering capability would be much better.

Thank you very much.

Regards,

Jan Laznicka

Connect currently provides lineage for a datasource at the table level but we would like to be able to have this at a more granular level i.e. the column level so if a specific column changes in the table (e.g. data type change), we are notified of the change and able to identify quickly assets that will be affected.

@SeanAdams @revathi

Hi Alteryx Team,

Can you please add PDF or Word Download Option for Glossary Definition Term in Alteryx Connect. 

Thanks

Sri

Deploying Connect on MS SQL Server (https://community.alteryx.com/t5/Alteryx-Connect-Knowledge-Base/Deploy-Alteryx-Connect-on-MS-SQL-Dat...) we would like to use integrated security option when we setup jdbc connection.

For now workaround looks like this:

1. Copy the file “sqljdbc_auth.dll” into Alteryx folder:

2. Inside SQL Server Management Studio add Active Directory user to logins folder:

3. grant him(or her) the rights to connect to database. For test purposes I`ve granted sysadmin role:

4. Edit the connect.properties file:

Here we don`t need user and password lines at all. We only need to add “integratedSecurity=true”

Because there will be used credentials of user, using which Alteryx Connect was started, not the user which was provided in these red records.

But we still need these two records, because Alteryx will give us an error if it will not see these records inside config file during startup.

That`s why these records still must exist inside connect.properties file.

5. Setup Alteryx Connect service to start Alteryx Connect with Active Directory user, for which we granted access in SQL Server:

6. Start Alteryx Connect service and check with Ms SQL Profiler the user which is performing requests to database:

It`s Active directory (windows) user.

We found out that we are able to setup Nexus modes in global config settings.

We can distinguish modes by relationship types and object types. It is nice.

We would like to distinguish Nexus setting by object’s location as well (to show/hide objects in diagram by their location).

For example to show terms from one folder and to hide terms from another.

We ask you to make possibility to setup Nexus to filter out objects by their location.

Establish lineage/relationship between SAP HANA and Tableau (reports)

Issue

By default in Connect, we are unable to see data lineage between SAP HANA and Tableau. The standard Alteryx loaders (i.e. Tableau & SAP HANA) in version 2019.2 are unable to detect the HANA object link due to a misaligned naming convention (“::” vs “/”). This specifically occurs when Tableau dashboards are pointed to HANA synonyms (“::”) as opposed to the views (“/”).  

Background

In SAP Hana, there are two types of objects – views (which are present in relevant schemas, e.g. “_SYS_BIC”) and synonyms. The synonyms are created for each view by default and always appear to be in the schema “PUBLIC”, however, this is not actually an explicit schema. According to SAP, public synonyms are the official access interface for the system views.

When loading the objects from SAP HANA, the object names contain the package name as well as the object name (view etc.). These two are differentiated with a “/” when loaded from the “_SYS_BIC” schema (e.g. “GSTP_VALIDATION/CA_01_C_PDA_GRA_001”), however synonyms from the “PUBLIC” schema are loaded with a differentiator of “::” (e.g. “GSTP_VALIDATION::CA_01_C_PDA_GRA_001”). An example below, where first row is the view and the second row provides the related synonym:

Thus, the lineage is not established when the SAP HANA objects from schemas are loaded (containing the “/”, and at the same time, the Tableau dashboards linked to the synonyms (“::”). This is because the object names are quite clearly different, and the loader doesn’t currently resolve the synonym to view translation (as per standard SAP HANA SQL parsing behaviour).

Loading problem analysis (using standard Alteryx loaders)

On SAP HANA side, when running the loader, user can (optionally) define schema name(s) they wish to load, however:

1. When schema is set as “_SYS_BIC”, the loader only loads the views from this particular schema, with no synonyms.
2. When schema is left as blank (i.e. no value entered), the loader loads everything, however in the “_SYS_BIC” schema, only synonyms are loaded and not the views.
3. When schema is set as “PUBLIC”, the loader does not return anything - fails as there is no corresponding schema in SAP HANA (it is only virtual space, not an explicit schema)

The SAP HANA instance in question is very, very large and due to volume limitations, as well as the potential risk of performance degradation, it is not feasible to load everything.

When we’ve examined the H2 staging tables, we find that Connect has loaded (therefore is technically aware of) the synonyms from the “PUBLIC” schema but they are not ever loaded beyond the staging tables. This is obviously required if we wish to represent these relationships within Connect.

Idea/Fix:

• SAP HANA loader – it seems there is a problem of loading synonyms from the “PUBLIC” schema and the loader should be adjusted in order to allow this.
• Tableau loader – needs to be able to refer to synonyms defined in a SAP HANA data source, meaning whenever there is a “::” in object name and the schema name is blank (i.e. empty string representing the “PUBLIC” schema), then it should be defaulted to “PUBLIC”. This logic would ensure that a schema name is always present, even in cases where the schema name is retrieved as blank (which then would become “PUBLIC”).
• In case the schema name is left blank for any reason, it should be populated as “PUBLIC”.
• @DanH suggested to customize the Tableau loader as a temporary fix to replace the “::” with “/”, which is certainly possible but comes with a risk of incompatibility with future releases. Also the customers strong preference is to deliver a solution which only utilises standard loaders.

In a large environment- especially an analytical environment - copies of data will often appear in multiple places.    an example of this is where a copy of a shared dimension or a shared piece of reference data is copied in multiple different data marts.

In order to manage this - we need to be able to mark these as copies of each other so that we can point folk to the golden-source; and so that we don't need to document this asset multiple different times.

Example:

- Client List appears in the data lake; on the Sales data mart; on the Finance data mart; etc

- We would want to group all 3 of these together; and mark the Data Lake version as the master; and all the others as copies.

User experience:

• When I navigate to any Sales assets - it tells me that the Client List is a data asset which is used
• When i click on this - it tells me that the sales version is a copy - and directs me to the one on the data lake

NOTE: There are circumstances where a copy may be deliberately filtered or incomplete (for example - regional subsets of clients) - in this case the relationship needs to be "Partial Copy" not "Copy"

CC: @DavidM @Arianna_Fuller