[Cdash] How to handle huge test output

Martin Apel martin.apel at simpack.de
Fri Dec 11 07:34:39 UTC 2009


Julien Jomier wrote:
> Hi Julien,
>   

>> we are indeed using MySQL as the database backend. As I don't know the
>> internal table organisation of CDash, I am wondering, if compressing
>> whole rows wouldn't slow down accesses
>> to the tables very much. Wouldn't it be better to only compress the
>> column containing the text output? This could be done in CDash upon
>> access of this column only.
>>     
>
> I played with the compression this morning and using zlib to compress
> the test output we can get up to 8 times compression. I think this is
> the way to go. I'm going to implement that soon.
>   

that would be really great. I expected that the compression ratio would
be rather high, as it is
pure ASCII (or even XML?) which is compressed.
>   
>> I already tried to put a feature request in the bug tracker, but the bug
>> tracker, which is linked from www.cdash.org does only contain category
>> entries for CMake, CTest etc., but not for CDash.
>> So I thought I'd ask on the mailing list first.
>>     
>
> There is a project CDash on www.cdash.org/Bug. We probably want to move
> the bug to the CDash project.
>
>   
Now I found the reason: I looked at the categories, but didn't notice,
that you have to switch the project, not the category.
I entered a feature request, it got the number 10027.

Best Regards,

Martin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/cdash/attachments/20091211/ea1f06d8/attachment-0002.htm>


More information about the CDash mailing list