Community Spring Cleaning week is here! Join your fellow Maveryx in digging through your old posts and marking comments on them as solved. Learn more here!
The Product Idea boards have gotten an update to better integrate them within our Product team's idea cycle! However this update does have a few unique behaviors, if you have any questions about them check out our FAQ.

Alteryx Server Ideas

Share your Server product ideas - we're listening!
Submitting an Idea?

Be sure to review our Idea Submission Guidelines for more information!

Submission Guidelines

Featured Ideas

Hello,

 

This issue relates to an inability to input data from a database where access to only certain columns is permitted. This is due to PII data being present in the data.

 

I am trying to pull data from one table at a time using the standard ‘Input Data’ tool. 

As I do not have table level access I am explicitly specifying the column names and not using a wildcard. Please see the example below, query highlighted in yellow and the returned error in red. You will see the error message returned from Alteryx suggests a * wildcard has been used despite specifying the exact fields to pull.

 

Several of the Ford GDIA team and Ford HPC team have reviewed this with me and cannot assist. The HPC team believe it is a bug in Alteryx which has been reported by other users internally, that is ‘Select *’ commands being sent from Alteryx despite specific columns being outlined in the query to the data lake. As we only have access to specific fields within Ranger (due to PII data) the select * fails and returns an error.

 

I can confirm my ODBC connections are all set correctly as I can pull from certain tables where I have full table level access but not others. Access rights are all in place as I can use the same query on Ambari without issue.

 

Thank you for any assistance you can provide,

Michael Higgins

 

Example Query

 

Select STATION_ID, STATION_DESC from dsc60082_qlscm_tz_db.qlsc_station limit 10

 

Info: Input Data (3): ODBC Driver version: 03.80

Error: Input Data (3): Error SQLPrepare: [Hortonworks][Hardy] (97) Error occurred while trying to get table schema from server. Error: [Hortonworks][Hardy] (80) Syntax or semantic analysis error thrown in server while executing query. Error message from server: Error while compiling statement: FAILED: HiveAccessControlException Permission denied: user [mhiggi37] does not have [SELECT] privilege on [dsc60082_qlscm_tz_db/qlsc_station/*]

 

---

 

Hi Michael,

I believe it would require turning off the queries we currently run to retrieve metadata and retrieving metadata only for the columns selected in the query.

Please submit to our Idea Center for consideration.  Thanks!

--
Angela Ogle | Customer Support Engineer

Would like to see a Notification hierarchy implemented to Gallery. Currently, the settings that control notifications are at the Gallery-level....where only the admin can control, for example, notifying users if a new workflow was added to a collection they belong too. Could this setting be inherited, but then, for example, a Collection owner can implement their own notification settings that overrides the Gallery default? Using the same example as before, perhaps the Collection owner could disable notifying their Users if a new workflow is loaded to the collection.

 

Nick

While working with Alteryx Server, I noticed that there is no way for an Premium Artisan to organize the workflow results. With processes that are running hourly, each day, the workflow results gets very cluttered with the same things, and it takes some time to clean up. It would be nice to add a search bar that would filter out workflow results dependent on what you are entering in the textbar. Below is a very lo fi mock up. The spot where I entered "dispositionReport" at the top is the text bar that filters the workflow results below it. The red lines are simulating the removal of said records. Then offer a select all option to then delete all of the workflow results records. 

 

workFlowExample.PNG

Insights should work like Apps. We should be able to publish insights then grant permission to other people/studios to access the visualizations. I have many clients / users that can benefit from having access to the Insights i'm gathering from Alteryx. 

0 Likes

As part of the Persistence Expiration processes, within Designer we have the ability to set a retention threshold of the results tab.   i.e. 30 days.

After enabling this - all completed results are purged  but all the "error" results remain.  And this depending on the original count can run into the thousands, such as in my environment.   Id like to see the "Error" results become part of the clean up processing because of the following reasons:

1.  We do not have dedicated admins that have time to manually or by group delete these error result items.

2.  Most if not all - errors are resolved immediately.   if there were to be kept as a reference, a screen shot of the results are normally taken and filed away.

 

Support says this is intentional for resolution tracking purposes - but to counter - as I stated in item 2 - most errors are immediately looked at and worked on. So there is no reason to keep errored results. Especially when they are time-stamped dates greater than the expiration values selected.

 

Hey Sever Gurus - 

 

There might times where it would really nice to have a group of workflows that were logically related that you wanted to run without having to invoke a 'run now' for each workflow individually.  Sure, you could technically get there with the Runner macro, but in my case, there are intermittent scenarios in which I'd like to be able to run a set of related workflows to kick the tires, but they run at different schedule frequencies during normal operations.  As it stands now, I have to go hunt through 39239 different workflows, find the 12 associated with project X and tell it to blast away individually.  This brings me a great sadness.  

 

So what I'd like to be able to have is some flavor of named / logical grouping I could attach to a workflow at upload to scheduling time, I could then tell the scheduler to show me / run everything attached to that grouping.  

 

Thanks!

 

brian

There are two key APIs that we use for extracting canvasses for governance & surveillance:

- the Workflows API which lists all canvasses on the server;

-  and the API call which returns a packaged version of the application by AppID

 

Unfortunately - when we query for all workflows to get the first list - there is no flag to indicate that a subset of these are actually deleted.   The result is that you need to make the second call to find this out which is wasteful.

 

Please add a "deleted" flag to the Workflows API so that we can identify which applications have been deleted without having to call the specific App ID post.

 

CC: @jalvarezv

We are taking an incremental feed of all server canvasses in order to look for patterns - however there is no option on the server admin APIs (the Workflow API specifically) that allow you to filter by date.

 

Please can you add to this API to allow API users to specify an upload/updated date filter and only take items after this date.

 

cc: @jalvarezv

0 Likes

Currently only 5 workflows are displayed per page in a collection.  Currently we have about 30 workflows (soon to be about 100) and paging through workflows to find the one you want to run is time consuming.

 

It would be great if there was an option on the page so the user could select the number of workflows per page.

@SteveA did an article about server logging a few years ago which was very helpful.

https://community.alteryx.com/t5/Engine-Works-Blog/A-deeper-dive-into-Server-logging/ba-p/22389

 

... and in the training events at Inspire this year there was a session on how to find and interpret server logs.

 

it would be very useful for large server environments to move away from discrete log files as much as possible, and instead move towards a world where the logs are stored in a data format, and one logging infra is used for all so that an admin team can see these all in one place.

 

Granted - the initial boot-up of a server, where it needs to connect to the logging DB is going to be a challenge - but after this initial primary boot sequence it would be great if all server logs and logging events were stored in a database of some kind so that we can analyze and collect events.     That way, server admins can have a great UI experience; as well as a great analytical experience without having to regex through multiple different .LOG files.

As Alteryx makes more and more use of the Public gallery - it would be valuable for Alteryx to explicitly have a section devoted to items that are certified by Alteryx so that we can be safe bringing them into our corporate env.

 

The assets created by Alteryx are currently listed with Alteryx as the author, however it would be useful to have a completely separate section with Alteryx assets only, and with these broken down by type.

0 Likes

If you are viewing schedules, on the Queue Panel, you should have the option to refresh the information to get an idea how far things are progressing.  If you swipe back and forth between panels you get updated % done information, so it is definitely available.  Not the most pressing need, but a very easy fix. 

 

Ninja Edit:  Did not know there was a dedicated server ideas forum so I removed a part. 

each canvas in our environment goes through initial dev; then a testing phase; then into production usage after completing testing & peer review.

 

Currently the server environment has no concept of progression or promotion so we have to set up our own Dev; UAT; and Prod gallery & server - this really is very clunky because assets have to be manually copied across.

 

What would be more effective is:

- Allow a single asset to have a lifecycle (like you do in GIT) - so that it can go through dev and testing as a branch of the main code

- Once ready - user then changes the lifecycle type to "testing" or "UAT" and pushes it to users

- Once signed off, the user then promotes this to a production flow, and it replaces the current production version

 

There are a few areas where we'd need to do some thinking (dependancy tracking; branching; changing connection details as you promote so that you can hit the prod data stores etc) - but fundamentally this would go a VERY long way towards eliminating much of the admin involved in running the gallery.

We've confirmed with @KevinP that the Alteryx Gallery / Server API does not currently support downloading canvasses in unzipped data-stream format. (https://community.alteryx.com/t5/Alteryx-Server-Discussions/Server-API-Download-packages-for-surveil...)

 

Please can we add an API end-point that allows the following:

  • Connect to the gallery and download a canvas; along with all of its sub-canvasses and helper macros. 
  • Format required:
    • Plain XML (not zipped)
    • Provided in a data stream so that we can perform analytics and surveillance on this data
    • Fields required
      • Canvas filename
      • Application ID
      • Some unique ID per macro / sub-macro
      • XML for the canvas
      • Last run date
      • Last changed date
      • Environment (this is important once Server actually allows a Dev; UAT; Prod progression cycle - so we can see which environment people are in)
      • User details who uploaded
      • User details who are marked as the owners
      • Any other meta-tags on the canvas (important once server allows for meta-tagging of canvasses for categorisation)

 

Current API only allows this to be downloaded to a zipped file; which then needs to be unzipped; and then loaded back into a data stream.

 

Thank you

S

 

CC: @revathi @AshwiniChezhiyan @LizaNemchynova 

 

Option to update ownership to any other valid license holder as needed.

 

Use Case: Currently the canvas owner is marked by default as the user publishing the canvas to production.

However, we have seen instances where the person moves teams and would like to hand it over to another person.

Enhance the USER tab in Gallery for showing the active user status, last login time

 

Easy to filter out the active user

Allow Admin to kill user session 

Hi there,

 

Currently, when we download a version of Alteryx - we need to download a very large number of binaries across many different download locations.    Once we get these in, they need to go through InfoSec audits; then be packed for deployment etc.     this creates a very large admin overhead in that folk need to check for updates across multiple binaries and also create a significant number of different installs which then need to be rolled out across hundreds of workstations.

 

The request would be to change the approach to downloads for enterprise customers:

- Admin logs into the download portal

- They then select the components which they need (Designer; predictive; CREW macros; Microsoft starter pack; Tableau starter pack; JIRA connector from the gallery; additional predictive tools from the gallery)

- this then becomes the defined selection set.

- When any of these items are updated - the admin is notified

- When they download - these are then all built into a single installer binary which we can then download and roll out to all our workstations.

 

Many thanks

Sean

 

cc: @dataMack @adrianloong

 

 

In order for us to manage the large number of canvasses on our server - we need to add the ability for Admin teams to require additional attributes on every canvas:

For us, these mandatory attributes would be:

- Which team do you belong to (dropdown)

- What business process does this serve (dropdown - multiselect)

- Primary & secondary canvas owner (validated kerberos)

 

For the ones that have dropdown lists - we can provide the master data into a drop location or into a manually configured list on the server.

0 Likes

I am running version x of Alteryx locally and my server is running x-2. If I publish a workflow with error reporting via email, the server is not able to recognize that my workflow has error logging because it automatically kicks a version error. The events portion of the XML should be read regardless of the server version and trigger an error when this happens.

There should be a way to set limits on Gallery so users are not able to set up to many scheduled apps to run. Currently you can either allow all users to add as many scheduled jobs to the run or none at all. If one user sets up whatever number to run, keeps other users from running their jobs. Either allow Admins to set limits or have server shift jobs around so one user cannot lock a gallery to only run his/her jobs. A combination of both a set limit per user and better job management by gallery should also solve the problem.