Hi Jason. It appears that your first view is pointing at a Tableau Server published datasource. What this means is that your workbook just has a pointer to the data that lives up on server and all the details are up on server. If you want to see the underlying tables that this published datasource points at, you can either download it from the server, or you can right-click on the datasource connection within the top left window of the Tableau Desktop and then "create local copy". I hope this helps.
Thanks for your response Jeff. That does work, but only temporarily. Up until today, I could open my workbooks and go to the Data Source tab and review my data source settings like shown in my screenshot (those shots are redacted to keep data private). As soon as I did what you recommended, closed the workbook (saved changes) and reopen it, I'm back to square one.
All of these workbooks have 'live' connections to our sql server.
Do you open it back up from a location saved on your hard drive or do you open it from Tableau Server? If it's the latter, square one makes sense as the workbook needs to be republished in order for it to take effect.
And while it may be a live connection to SQL Server, if you are pointing at a Tableau Server published datasource (denoted by the icon that looks like: ) then Tableau Server is acting as a proxy to your SQL Server for the live connection.
I open it back up from my 'workbooks' directory on my hard drive. I have also tried republishing it to server in my 'test' project with no luck. I'm trying to understand what changed when I opened it this time to work on it vs all the other times. It's the 1st time I've run into this issue. When I open the other workbook which is connected to the same db I get my original view.
I do notice when I used your method above and opened the datasource it shows that it connects directly to my sql server. When I click on the 'data' menu at the top, the previous view shows the sql server name with a check mark at the bottom. The current view does not show anything like that.
Does it have something to do with the workbook type I'm saving it as? Packaged vs non-packaged? [EDIT: It does not, I just tried that also]
If I open the workbook, then click on the Tableau button and then click on the saved data source, it appears as usual. However, as soon as I change to the dashboard tab and come back to the data source tab it reverts again back to this other view.
Ok, this may help some. I couldn't understand why it was connecting to Tableau Server at all (and still do not). I stopped the services on the server and it will not connect (as expected). Since my workbooks are housed on my local machine, why is it trying to connect me to Server? My other workbook opens directly from my PC and to the SQL server automatically. I need to figure out what changed in the other workbook that it is not available "offline".
Thanks for your help Jeff, you've got me troubleshooting in the right direction.
I went into Publish workbook to review what was being published as I noticed each time I connected to the datasource via the Tableau menu, it was incrementing the source name. There should be only one. Now I
need to see if I can figure out how to remove all the unwanted datasources. I believe it's the last one that is also causing my problems.
Based on the description the only thing I can think of is that there are two workbooks in two different versions in two different locations, one lives on Tableau Server and one doesn't, and one is being opened and confused for the other. Here's an example from my own workspace, there are two "DailyEncounters" workbooks in the list, one is local, the other on Tableau Server:
If the workbook names are long enough the [tabdatablick... isn't even visible so the two entries look the same, even though they are completely different workbooks stored in different places. This happened to me more than once when I began publishing workbooks & data sources to Tableau Server. This can lead to version control nightmares. For example I if I open up the Tableau Server instance of the workbook, then save the workbook to a local file (instead of using the Server->Publish menu), then open up the workbook from Tableau Server again I'll see an unchanged workbook because I didn't republish it, this behavior fits the description as far as I can tell.
In my case the solution was that I grew to trust Tableau Server's backups enough that I stopped keeping local copies of workbooks and used Tableau Server as a "source of truth" for my workbooks and data sources. So workbooks lived locally on my hard drive, or were published, but not both.
If it's not this, then I'm out of ideas and would suggest contacting Tableau technical support.
Thanks Jonathan. I try to be careful not to do what you say (I use the mouse hover feature) to make sure I'm grabbing the correct file, but it's entirely possibly I messed this one up. Are you saying it's a 'best practice' to just keep the one copy on Server and edit from there?
Also, how do I delete those extra data connections I showed in my screenshot above? It will not let me remove them from the data tab, I can only 'close' them.
We need to be clear on whether you want to publish the *workbook* or re-publish the published data source. What's happening here is that you're downloading local copies of the published data source into your existing workbook then publishing the workbook (with all those local copies).
When I'm working in a Tableau Server instance that has published data sources and workbooks I've found I need to be very clear about the task I'm trying to do. Am I:
a) editing the workbook?
b) editing local calculated fields in the workbook?
c) editing the published data source itself (and republishing it)?
For a & b we just edit in the workbook and then re-publish it. For c (editing the published data source itself and republishing it) the process I use and suggest is:
1) open a workbook (I usually do this in a completely new workbook so I don't confuse myself)
2) download the local copy of the data source
3) make edits to the data source
4) right-click on the data source in the Data window and choose Publish Data Soruce
5) make sure the data source has the same name & project information as the source you want to overwrite, then publish it.
a) If you *don't* get a prompt about overwriting an existing source then something about step 5 got screwed up and now you have two different published data sources, the original one you downloaded in step 2 and the new one from step 5. You'll need to re-publish with the correct name/location and then go on Tableau Server to delete the accidental version.
Closing them is how to get rid of them.
re: "best practice" to keep one copy on Server. I don't know the health of your Tableau Server environment so I won't say "definitely just use Server", if it's trustworthy then that's what it's designed to support (mostly***). Tableau even started versioning workbooks with Tableau Server v9.x so you can track versions there and reduce the need for local copies.
*** The mostly part is that when we're using published workbooks and published data sources there's no way to keep versions in sync, if I I'm going to make sufficient changes to the published data source that the existing workbook(s) would break then I have to be ready to publish up new versions of the workbooks. The general workaround is to have separate dev/test/production environments but even within those we can run into this problem.
Also one little practice that I use is that when for whatever reason I really do want to have a local copy of a workbook I use the File->Export Packaged Workbook instead of Save As... because that keeps the exported workbook out of the recently opened workbooks workspace (and helps me avoid the problem of accidentally opening up the wrong version), and I'll make sure to add a date/version stamp like myWorkbook 20161206 v1.twbx.
Great, thank you! This is all great information for me. Unfortunately, I'm the only resource here for Tableau and I'm new, it's new, etc etc so I do not have anyone to troubleshoot with or bounce ideas off of. I greatly appreciate the help from both of you.