Version 3.22.17+200525

Not sure if this is the correct place for this question.

The Submit button only reloads the survey and capture it as incomplete. No data is collected then.

Number of questions/groups: 266/16
Survey logic file shows no errors
PHP version 7.4
Max_input_vars 50000. Even increased it to 250000 but still no resolve.

All other surveys are working fine. This is the only one that do not submit.

Survey attached.

Any suggestions?

Wow, there are 16 groups with a total of 266 questions. Not sure if I will have time to go through this and test it.

You have over 1000 subquestions.

My guess would be that you would run in the limitations of the database, but then you actually shouldn't be able to activate the survey.

Hi holch,
no definitely not would I expect that you go through the survey. I merely attached it as reference. I thought I might have missed a setting in PHP but all seems fine there.

I also suspect that the survey has to many questions at first. I followed some suggestions that I delete group by group to test if it will submit. Eventually, I found that section 7 is the culprit. Still not sure what exactly in that section is stopping the survey from submitting. I am now systematically removing all the conditions in that group to see where the problem lies.

Thank you for the reply. Appreciated. After using LimeSurvey for a few years a simple issue sometimes makes you scratch the brain.

no definitely not would I expect that you go through the survey.

As you have probably seen that I have "weakend" this a bit, after reading it a second time.

Actually, to really find the problem it probably is necessary to go through everything, otherwise it is difficult to reproduce the problem.

Eventually, I found that section 7 is the culprit.

That is already something we can work with, I think. But how did you determine that section 7 is the problem? If the assumption is "too many columns", than it might not be section 7 itself, but just that section 7 has enough questions, so if you delete it, it is "enough" to work.

I will have a look later, if I find the time.

I'm here to learn. Never to old for that.

Sometimes a fresh approach or a different method is the key to solving the problem.

I started by deleting group by group from the bottom. Testing the survey after each group removed. When I removed section 7 (sorry, see now it is group 9) it submitted. I then imported all the groups I removed except for group 9 (section 7) and it submitted.

I now have placed group 9 (section 7) back, deleted all the conditions in the group and removed all filtering. Still no luck. Next step will be to remove group 9 (section 7) question by question (from Q123 up to Q103). Same method I applied when removing group by group. This should show me which question in group 9 (section 7) is responsible for my headache.

Thank you for the time you spent on your replies to me. I realize time is a valuable commodity in this world.

This is a very good and systematic approach, makes total sense. Keep us updated and once I find a little bit more time than just a few minutes, I'll try to have a look based on your latest status.

But I am quite confident, that you will be able to drill down to the problem with your approach.

What is your DB system ?

If you use MariaDB, Mysql or pgsql : an activated survey must be OK

But : i have bad experience with MSSQL (and a lot of columns) : activate seems OK. Survey have a lot of array text : ~800/1000 columns of data.

If user put only 2/3 letters in each data : it's OK, but if there are more : this broke all save and make empty response … really awfull.

I must report it …

Hi Denis,

it is MySQL and at the end part of the problem (as holch has suspected due to the size of the survey).

In the beginning I suspected one of the groups in the survey to be the issue. After I removed the specific group, the survey submitted without issues. I then created a new survey with only that single group and tested it. It too submitted without issues. The suspect group was re-inserted into the main survey and by removing question by question from this group I eventually found that the survey submit if I remove the last 5 questions. The penny dropped when I removed a question from another group and added a question to the original group. A simple one for one swap. The survey submitted without issues then. This pointed me to the size of the survey. I redesigned some question groups and although the survey is still big (15 groups, 258 questions, 1047 subquestions) it is working now.

Big thank you to you and holch for pointing me in the right direction to solve this issue.

Hi,
even if the issue is solved,
here my findings.

When I imported your survey, code of question Q58 was renamed.
So I investigated the lss code.

And here
<title><![CDATA[S4Q34]]></title>
<question><![CDATA[<p>Q58.    To what extent do you agree or disagree with the following statements about cryopreservation using the same totally agree, totally disagree scale.<br />

This because the same code was used in Q34.
<title><![CDATA[S4Q34]]></title>
<question><![CDATA[<p>Q34.    To what extent do you agree or disagree with the following statements about stem cell harvesting, research or therapy using the same totally agree, totally disagree scale.<br />



Some things to consider:
Even if you use LS as a CATI tool, it is not advisable to display "all on one page".
You see, you have an intro text in each section. Here you could have splitted the survey.
And why didn't you use quotas to screen out people?
S7 could have been calculated automatically; only in case of two equal percentages ask "Which one?"
Ease the life of interviewers.

Joffm