Interesting SAS Viya4 bug with Excel exporting

[Berux]

Hi,
I have an unpleasant problem with viya4.
I have created myself a report in VIYA4 Visual Analytics containing a multidimensional table type object.
Unfortunately, when exporting the multidimensional table to excel, I get an error 500 as shown in the screenshot below. Interestingly when I selected such values in the filters to which the table was practically empty it generated excel without any problem. The problems occurred when some values started to appear in the table.
Interestingly I created 10 reports and in 9/10 completely this problem does not occur. Multidimensional tables often contain a lot of data, calculated elements, etc. and yet they generate beautifully except for this one, which does not generate even with the minimum amount of data
Please help and best regards

[TSBruce]

For this particular issue, I would suggest starting a Technical Support case so that logs and other diagnostic information can be requested.

[Berux]

We have send a couple mails to SAS Support and we didn't get an answer to this day.

[Madelyn_SAS]

Which email address are you sending your request to? Ensure that you are using support@sas.com

You can also use the customer portal. Go to https://www.sas.com/en_us/contact/technical-support.html to see the four methods available: Customer Portal, Chat, Email, Phone. 

[Patrick]

---

@Berux wrote:
We have send a couple mails to SAS Support and we didn't get an answer to this day.

---

Whenever you raise a track with SAS Tech Support you will receive an automated confirmation email pretty quickly. If you haven't seen such a response email then make sure you've sent your question to the right address and also verify that the SAS TS response didn't get moved to your spam/junk email folder.

[Tom]

I suspect that there is an issue with some of the data in the report that is causing the EXPORT to fail.  You can use the methods you used to test to isolate a small set of observations that cause the trouble.  I would recommend a binary search.  So find some way to divide the data into two groups.  Test each subgroup. 

If both succeed then the issue is probably total size.

If one of them fails and the other succeeds then concentrate on the one that failed.  If both failed then just pick one.

Repeat until you either find that all subgroups succeed (again that looks like a size issue) or isolate one or a handful of observations that cause the problem.

Now you could try eliminating variables and see if you can find which variable has the strange value.