Seeking advice: how have YOU positioned Server in your networks ?

Question

Hi !

We've just acquired our first instance of Server with the idea of moving off our Scheduler instances.  But we're running into challenges as we try to position Server in our landscape.  We are on O365, using MFA and we are progressively pushing all our stuff to the cloud.  We are trying to use Sharepoint as our file repositories (we use OneDrive sync for our desktops, My Documents etc and for Teams collaboration).  This is causing us problems when our Server-run workflows try to access Sharepoint-hosted file locations to read/write csv etc.  We've been getting away with OneDrive sync on our scheduler-run workflows, but would like something more robust on Server.

So, how are YOU positioning your Server instances ?  Where do they read/write files to - traditional file servers ?  Sharepoint ?  Azure, Amazon or other cloud locations ?  Are you using service accounts on your host machines ?  And what about MFA ?

I would very much appreciate your feedback on this & any advice you may have.  Please also share how many instances of Server you have.

Thanks in advance for your help.

Cheers,

8088.

jensroy · Answer

Hello!

Well Sharepoint Online = Onedrive, and I don't think Alteryx has support for Onedrive (though there is a connector in the Gallery for that).

So we are using sharepoint 2016 on premis for excel files. Here, we still have authentication issues as the AD token expires and Alteryx seems unable to pass on those credentials.

So I made a macro where the user inputs filepath and sheet name, then I convert that into an UNC path, then I have alteryx create a bat file saying "Open this folder with File Explorer" then waith 10 seconds, then open the file that the user specified". This way, Windows, not alteryx, is opening the sharepoint folder and the access token works fine.

Next we also use sharepoint lists. This works really well, except the size issue. Only 5000 records per list i think. but it works well for many cases, giving the users a way of maintaining small data sheets.

It is also possible to build a REST API to get the files from Sharepoint Online, but I did not open that door yet.

But most of our data is based on our datawarehouse(es). Be it oracle, SAP or MS SQL.

And alteryx's role is most often as an ETL tool between the DW main ETL jobs and reporting. Meaning that alteryx prepare datasets for our BI platform, based on DW data and sharepoint.

Other ways to deal with excel files is to read them in to the datawarehouse using Alteryx. Here you can either develop an Analytical App in the gallery where you give users the option to upload files, and then have alteryx commit those files to the DW, or you can have local workflows doing the same.

Hope this is useful feedback on your case :)

CristianoJ · Answer

Hello,

I am actually responding from support.  Having a file server or SharePoint file repositories is a matter of preference, and we have seen both.

Please go ahead and send the details of your particular scenario to support@alteryx.com and we will be glad to assist you.

8088 · Answer

Cristiano,

Thanks for reaching out.  Although we do have a specific issue with our Server deployment that we are still trying to bottom out (have been working with Andy Main) and for which we would gladly take insight, my post was more general in nature.  I'm trying to understand how folks "typically" deploy server so that server-run workflows can read/write csv files.  Do folks use a regular file server for that purpose ?  Do folks have success using Sharepoint file repositories in that way ? That's what my post was trying to solicit from users who have experience in this area.

Regarding our specific issue about not being able to have Server flows interact with Sharepoint file stores using fully formed UNC file paths may be linked to our use of O365 cloud services and MFA ... but that's a guess.  I'd be happy to raise a support case with the specific details of our config.