How I wish upon a Tableau star.

How have times changed. I sound like an old man. I mean I’m 37 years old and have been working in the Data industry or around data and databases for pretty much my entire professional life. Its pushing on 17 years. Long gone are the days of writing out your code and you leave it to run overnight. Walking out the door with fingers crossed. Only to come back into the office the next morning with an error.

Query failed.

No debugging message. No helpful or sometimes not so helpful error code which you could Google an answer for. I mean I have massive respect for the really old school guys. They were literally pioneers in the data world.

Now we have so much power and resource at our fingertips. Instant answers to (in my case) badly written code. Not to mention software companies who set themselves goals (and deliver) at quarterly release cycles.

I’ve used vendors in the past where you contact support with an issue. “I can’t do this”. Oh well yes it should be resolved in our next version. When is that coming? Umm, looks like 15 months from now…

There is always a huge buzz and excitement when Francois Ajenstat and his Dev’s come on stage at TC events. It’s one of the most anticipated sessions at TC. We get to see live demo of new features which are either going out in beta or very close to. So much great stuff that we are teased with.

There is still a few negative normans who say, well they demo’d that last year and it is still only just in beta. I mean software development isn’t exactly easy. Otherwise everyone would be doing it.

So enough of my rambling. I really look forward to these updates and beta programmes. Lots of my stakeholders and Tableau champions do as well. It is one of the best feelings when you can blow someone away with a new feature coming. Or even some really basic ones that have been missing for such a long time (transparent sheets).

My main point I want to make here is that in the past Tableau have not exactly got the balance right. Call me selfish but there seems to be a heavy skew towards desktop rather than server. As a server admin I see so many great ideas from the Server Admin community which we are still waiting for. I kind of hope that now Desktop is in such a strong place, some more attention can be paid to Server. I am just aware of how important the pivot to Tableau Prep is and hope it doesn’t detract from any Server development in the pipeline.

So, my Server wish list. Well some of these already exist on the community forum. If you agree, please give it an up vote. I’ve also added my other features which I think are missing and could make such a huge difference when managing Tableau Server for your company.

1. Allowing child projects to be locked, while parent is unlocked — newly named Zen Master, Mark Wu raised this in Feb, 2018. For me this plays a big part of my current project with migrating content from over 80 sites down to 2. I have many projects and sub projects which have allowed me to start this project, but managing permissions is a nightmare. I have to unlock the parent project to have the children different permission levels. But if I need to add a new active directory group to a child project or all the sub projects I need to change the permissions at all the projects and then select all the worksheets and add the new group to all these as well. It is hard to keep track across 3 levels of child projects. Up vote here: https://community.tableau.com/ideas/8455
2. I searched the current ideas on the community forum and wasn’t able to find anything that exactly matches my idea. If anyone else has seen something similar, please let me know and the ideas can be marked as duplicate. My idea is to create a two step process for publishing data sources. I have an issue with all my users with these permissions just publishing data sources. This could be ignorance which I need to address with education or just not realising that a similar data source already exists. Simply put a user can still publish a data source, but a simple approval from the server administrator/s would enable the data source to be seen if its verified and approved. By default the data source would be hidden from other users until approved. Up vote here: https://community.tableau.com/ideas/9659
3. This one is a must! Adam Freeman created this idea about allowing Server Administrators to view as other users. This would save so much back and forth when trying to set permissions correctly for projects. It’s embarrassing asking an end user “can you see now?”, “what about now?”, “ok, try now”. To be able to flip into the user you are trying to check permissions ensures data security on server. It can already be done within Desktop, so I really hope this should become a feature on Server. I can’t believe how long this has been an idea for (since 2016) and still nothing. Up vote here: https://community.tableau.com/ideas/6100
4. I tried to find another idea similar to this but wasn’t able to. It’s probably not going to impact you unless you use Google data sources. This is pretty relevant to me as we have just migrated all data sources away from Amazon Redshift to Google Bigquery as our primary database. Previously if I opened up a workbook I would need to authenticate once with Redshift. If I had 5 or 10 data source connections all from Redshift I would still only need to authenticate once. Now with Google BQ or sheets, the user has to authenticate each and every time depending on how many data sources are contained within the workbook. You are also able to edit the user and password credentials for Amazon Redshift data sources/extracts when published on server. Google data sources does not allow this. This means you can’t change the credentials over to a service account for billing purposes. It also causes issues if someone is away on holidays and their colleague needs to refresh the extract if it fails. Before using Redshift someone could just ask for the workbook to be transferred to them and then edit credentials — refresh extract now. Now using Google they have to download the workbook, authenticate for each and every data source. Republish, making sure that their credentials are embedded and then refresh extracts. Only for the actual owner to have to repeat the process once they return from absence. Talk about menial tasks! Up vote here: https://community.tableau.com/ideas/9660
5. Last one wish list server idea. So this was pretty much the focus of the Server admin meet up at TC18. Extensions gallery. Main issue is that by default extensions are enabled, meaning users can bring into a workbook and publish. Fine if you are very liberal with your server and data. Not so fine when running a deployment in a corporate sense. As Zen Master Tamas Foldi explained in his presentation to the Server community once an extension makes it into the gallery there seems to be no subsequent checks in place and anyone could be inserting all kinds of nasties into the code. He demonstrated how an extension in the hands of (in his own words) “a trustworthy guy” could extract all the data from within the Tableau workbook and send to another location. Pretty scary stuff. Not much is mentioned about this, but another Server guru and Zen Master, Paul Banoub highlights exactly this in his blog post. Always on the ball, another hot shot Server admin, and fellow Aussie, Mark Kernke created the idea to ensure that extensions were by default disabled on Server. At least this gives us a bit of security and piece of mind when upgrading and not suddenly shocked with a tap on the should from the infosec teams. Up vote here: https://community.tableau.com/ideas/9347