Tableau Hyper Extract as Input or Output Source

[anm252]

Hi all--I have a dashboard that has probably around 700,000 records that is published to Tableau online. The data source is a Tableau Hyper extract that comes from an Alteryx workflow that runs once daily as its output. This hyper extract is the input data source for many of my Alteryx workflows and I'm increasingly running into errors (that involve that the Tableau hyper extract) like this: "error opening database....is locked by another process...:operation would block Context." This hyper extract sits in a shared drive but has never been opened by anyone else or myself. Does anyone have any idea what is causing this and if there is a work around? This used to be an excel spreadsheet but the lag for the dashboard online was horrific.... having a Tableau extract as the data source has been so much faster.

[DataNath]

Unless I'm mistaken, the reason it's locked is because you are using it as the datasource in Tableau and therefore the dashboard is (silently) always referencing it. I've not used it personally but I know there's an 'Output to Tableau Server' tool for Alteryx. I'm not sure if using online and not hosting it yourself will affect use of this but if not, you could perhaps use Alteryx to update a published data source and switch your dashboard connection to a live one.

[anm252]

@DataNath  that actually makes a lot of sense and probably the reason why this is happening. I have tried the 'Output to Tableau Server' tool and had some issues with it (can't remember what exactly) but I may try going that route again. I know Tableau cannot use an Alteryx DB as its data source but is there anything in your experience that's anywhere close to as efficient (in terms of speed) as a hyper extract? Thank you for the quick reply!

[DataNath]

As .hyper files are specifically optimised for Tableau, nothing else will really be as quick for dashboard performance. Can I ask what the other workflows do that use the .hyper as an input? You may not need to use this, depending.

[anm252]

Hi yes--so the .hyper input is a large database of provider information that can change day to day, and usually the other workflows I'm using that involve the .hyper file are joined with various other financial files (in csv format usually) to produce forecasts and what not.

[DataNath · Accepted solution]

Ok, great. My thought in that case is that, in the initial workflow that creates the hyper file (that the dashboard runs off), create a second output of the exact same data as whatever filetype you like (.yxdb most likely as that's optimised for your other workflows). In your other workflows, point your Input Data to this instead. As this second output won't constantly be connected to something, you should lose the errors that you were previously getting by trying to bring in the Tableau extract.

[anm252]

I don't know why I didn't think of this. Thank you SO much!