[solved] memory exhaustion problem with repeated groups

[mudshark79]

Hey There !
I'm in the middle of finalizing one of my projects and ran into a rather unexpected problem. I created a list with several repeated groups. At this point I have a setup with 4 repeated groups, but I was planning to integrate up to maybe 15 groups. Each of this groups contains only little number of elements, but each row in the list will contain up to 30 repeats of those groups.

The list isn't supposed to be filled via a form but via automatism and is simply used for representation of data that can be examined and also exported and visualized.

After having finished only a small part of the endeavor it happens, that the whole list becomes slow while loading and memory exhaustion occurs. It makes no difference if a lot of entries are in the list but even 2 entries are sufficient to make the exhaustion occur. Unpublishing some groups makes the Exhaustion go away and after switching on debugging I see one huge memory Footprint in the Runtime, see attachment.

Fabrik is not newest version but maybe some weeks old already. What is the next step to get some more information about what is the root of this occurrence, if there is any?

Regards,
Matthias

 

[cheesegrits]

Does that group contain any CDD's?

-- hugh

 

[cheesegrits]

Hmmm, probably not CDD's.

Can you turn join merging off, see if that changes the behavior.

-- hugh

 

[mudshark79]

Oh yes, merging is the key. There are no CCDs, no Calcs. At first I had DB-Joins involved in the Repeated-Groups and turned that off in favor to simple Fields which didn't make a difference. Merging makes all the difference.

Is there something One can do about it? Maybe I will add a calc or php in the background for Compiling "Nice to look at" Fields anyway.... so I can use that fields for the list representation and will hide them repeated groups in the list... Even though it would be much better if this would just work as intended and as i think a moment this will bring other constraints.

Like already stated I'm not on the newest Codebase and as this is already live in some in on instance I'm hesitating to simply to an update because I'm afraid that something jumps out of the woodwork while I got other things to do. But I could try with a copy of course or to be more precise will do during the weekend for sure.
Sorry for not having done this before posting .

Regards,

matthias

 

[troester]

There's an option "Add Distict" in list Advanced settings.
You can try to set it to No.

 

[mudshark79]

"Add Distinct" Off doesn't make a difference....

 

[mudshark79]

Found some time to look at this again and got it sorted out, finally. I guess my approach was simply a bit stupid, like I already supposed. Why? The repeated groups do a lot of joining to other lists - this was clear. Bit It was not clear to me in the first place, that this join statements in the SQL spit out a row for every possible combination there is . Not even half way done I already had ~27.000 lines to break down on when merging for display. So that's why it took some time , I guess .
Anyway thanks for the Head-Ups.

Regards,

Matthias