nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

cross-tab row edge exceeded?

osuwariboy

Hi, I'm curious about something. I'm currently designing a report with a couple of cross-tab in them and they work without any problem, except when I start to increase the number of months to display (and by 'increase' I mean trying to display 3 months' worth of data). I then get this message:

"The result set length on row edge exceeds size limit 50"

Am I to understand a cross-tab can never display more than 50 lines at a time? That's a rather low value. And of course, it seems like it might be changed somewhere, but there's absolutely no signs of this property anywhere in Eclipse. And I've run the resultset in MySQL and it only returns like 100 lines, which is far from being a lot.

Anyway, can anyone tell me what I can do to avoid this error? Thanks in advance

Osu

Find more posts tagged with

Comments

mwilliams

What is your BIRT version? I have never seen this error. Do you create your reports with the designer or with code?

mwilliams

Take a look at this page in the doc:

http://help.eclipse.org/indigo/index.jsp?topic=/org.eclipse.birt.doc.isv/data/api/org/eclipse/birt/data/engine/api/DataEngine.html

osuwariboy

I'm using Eclipse Indigo which comes with version 3.2.23 of the BIRT engine. So I'm not really coding anything. As for the page, you suggested, I did indeed saw it after posting my question, but since I'm using Eclipse to design my report, I fail to see how I can actually change this constant's value (which name's comes complete with a nice typo he he he).

In any case, my datacube has the "row" dimension setup as such:

ClientDivision-->Option-->CoverageType

The thing is that the Option level is not always there for certain clients. When I select a relatively short period of time, the cross-tab comes out as expected, with the second level coming out as empty. But as soon as I expand and I have a client that sometimes have the Option level and sometimes not, this is where the error comes out. If it always has the second option level the error doesn't show up. You really have to have a case of a "sometimes" missing second level.

In any case, it's not really a critical for me, but you have to admit that's a rather odd error. Still, how exactly am I supposed to use the information in the link you gave me in Eclipse?

Thanks in advance

Osu

mwilliams

I was just pointing out where the setting is, in case you were coding your reports and happened to set the value. Is there any way that you can recreate this issue with the sample database? Or supply a dataSource, like a csv or xml file of data, and a report that uses it that has the issue? That'd be very helpful for testing.

osuwariboy

Hi, sorry about the long delay in response... but work's a nightmare these days. I've done a quick extract from our database and obfuscated the data. Though the PrimeSanteAdherent numbers are all fake, of course (I DO value my head he he he) they are all in the same range as the ones I had. Sadly, I don't have time to go through all the process of creating a stored procedure and a new BIRT report to test if it's really gonna crash, but I do know that this was the data that caused my report to generate the error. I've also included a screenshot of how my crosstab is constructed so you can try and replicate the problem.

So anyway, if you need more information, I'll be happy to oblige... and if you ever figure out what went wrong I'm really interested in knowing.

I hope it'll be enough to test things out and if not, I'll try and find time to get a BIRT report that does crash,

Osu

mwilliams

Sorry for the delay, myself. Do I have all the columns needed to create this cube? In the attached xls file, which fields correspond to the ones in the crosstab? If you can tell me those, that will help. If you can use this sample data and create a report to send me, with it set up how it should be, that'd be even better.

Thanks!

osuwariboy

On that front, I've made a bit of headway and managed to find where in Eclipse to set the parameter mentioned in the error message. I've included an image that shows where you can find it (default values are 50 for both fields, I've increased them to 100 and the message went away). Still, even though I was able to get my report to work, it still doesn't explain why he thought there were 50 of something, whatever that might have been.

So, concerning my data extract, you do indeed have everything you need to replicate the issue (I even think you might have a few extra columns that aren't used). Here's the mapping you'll need to construct the datacube:

For the column dimension, it's column C in my data extract, the DateDebutCouverture field.

For the row dimension:

Division would be column E
OptionCouverture would be column L
TypeCouverture would be column K

For the measure, the column used is column G with a simple sum. Also, I don't think you'll need the five grand totals at the end to replicate the bug, just the normal total will do.

So there you have it, I hope this helps.

Osu

mwilliams

I was able to run the report with no issues whatsoever, without changing the value from 50 to 100. Maybe it's the fact that I used the xls file as a source, rather than using MySQL, but I don't see why that'd matter. I ran it in the viewer, in preview, and in PDF with no issue. I guess another difference is that I ran the report in BIRT 3.7.0 because that's the version I have the XLS ODA installed in and it appears you're using 3.7.2. Maybe it's an issue in 3.7.2 only, but I doubt it. I guess if it works with you changing the value to 100, then you should be ok. Still weird, though. Let me know if you run into any more issues with this.