Actually the point here when you save a text file in your computer with file name "A" and another name with "a". That means now you have two files. It didn't overwrite the existing file . so the same thing is happening here also. Now the thing is your refreshing some other file rather than the file you want to do .
Rahul, I feel while refreshing the extract you'll be using Tableau's Command Line Utility. If yes then you need to make sure that the case is same as it is for the file else the server will not identify the files. Its not only true with files but with dashboards and views as well also if you are passing any variables they are also case senesitive. tabcmd or tabadmin commands are also case sensitive to some extent. Make sure the cases are correct else you'll be thrown an error.
Hi Tshasank Da,
If you do the same exercise in Windows server 2012 or Win 8.1 Pro, it will give you an error, implying that it is no longer case sensitive (as it was in Win 7/Server 2008).
In my case there is a process which Archives older files before the new files come in. So the older files are out and the new files come in with wrong case in the name.
Hence my question for the source file.
I am anyways planning to test it with a dummy file. Will let you know what I find.
Nope. Am not using tabcmd. I am using tableau Server schedules. Tabcmd works on HTML based characters hence is case sensitive. But I am not using it.
I have a tde which is published to server which sources data from a local file.
Since the data team has changed the case of the filename as mentioned in my original post, I wanted to know if it is case sensitive for file names as the next refresh will search for the path and file name that the tde was created with, no?
Yes it would be searching for the previous file name only. Because the system wont know that file is updated with other name. So it will throw you an error.
Hi Tshasank Da,
I tested the same.
Tableau Server is not case sensitive.
Edit: for file names for Sources which are extracted.
I created a file "source.txt". Extracted it and published it to server.
Then renamed the file to "Source.txt". Added a row to it and ran incremental refresh.
It added the row successfully to the source.
Unfortunately since I am in a secure environment, I am not able to send images to you show you the same.
check your server once you will have two connections in your server. with two file names
I did not get you.
Do you want me to check if there are 2 sources in Server?
Or do you want me to create 2 separate sources in Server with the same condition?
Or Do you want me to create a source created from 2 source files?
No You got me wrong.
I created a source using text File called "source.txt".
The data source name is "Test". I published "Test" to server.
Then renamed the source file to "Source.txt".
Then went to server and refreshed "Test".
This is what I did.
See that means you are changing the name for the text file. But the name of the published data source is same as "Test". From the starting we are discussing about this point if you change the published data source then you will get an error.
This is fine when you are doing an extract it is working . for example it is in live connection then you changed the source file name then it will work ?
First I am not working on a live connection. As I mentioned I was working on a extracted version of source.
So I wanted to know if the case of source file name matters or not. Now that I have tested it; I am safe even if the data team changes the source file name to upper/lower case.
However, like you mentioned in live connection I believe it will be case sensitive as the source will have an open connection to the file. If there is no file, it will display only the old data.