[mapserver-users] cgi-error: End of script output before headers mapserv

[Johannes Paul]

Hi Brian,
I'm not usinage fcgi on my side, and it is not related to file permission
or bad cgi script.
I've tried to work on my Osm postgis tables, and made the geom valid, like
Jukka said, but it didn't change anything...
I'm thinking it might be related with Apache MPM prefork module and the
MaxRequestWorker ?
Keep in touch


Le jeu. 23 janv. 2020 à 20:27, Brian <mulcahy.brianp at gmail.com> a écrit :

> So I have this error as well. I mostly see it when a connection times out.
> Like you mapserver shows no errors but apache has the script ended before
> headers error. Do you have fcgi enabled? If I disable fcgi I don't get that
> error, so It could be related to connection a user has to mapserver. Maybe
> FcgidMaxRequestsPerProcess, FcgidMinProcessesPerClass,
> FcgidMaxProcessesPerClass  need to be tweaked?
>
> If you do a google search on this error most of the responses says it is
> either due to a bad cgi script or bad file/dir permissions which in our
> case I doubt that is the reason. If you figure it out let me know.
>
> On Wed, Jan 22, 2020 at 12:31 PM Rahkonen Jukka (MML) <
> jukka.rahkonen at maanmittauslaitos.fi> wrote:
>
>> Hi,
>>
>>
>>
>> I would suspect data, especially because it seems to be OSM data. Perhaps
>> some invalid polygon again.
>>
>>
>>
>> You use &layers=default and I suppose that it means all the layers from
>> your mapfile together. In that case I would try to catch the failing layer
>> by adding layers one by one and in groups (first &layers=layer1, then
>> &layers=layer2, then &layers=layer1,layer2).
>>
>>
>>
>> Hopefully that way you can find the failing layer, and then because you
>> know the BBOX you may find the nasty geometry from the database. Without
>> your data and mapfile it is rather useless to make very much guesswork, but
>> a small change in bbox may select/not select a bad geometry, of change of
>> bbox may change scale so much that source data come from a different table
>> if you have a scale dependent layer group in the mapfile etc.
>>
>>
>>
>> If the issue is due to bad geometry and you use PostGIS you may find it
>> with ST_IsValid and ST_IsValidReason, and correct with ST_MakeValid.
>>
>>
>>
>> -Jukka Rahkonen-
>>
>>
>>
>> *Lähettäjä:* mapserver-users <mapserver-users-bounces at lists.osgeo.org> *Puolesta
>> *Johannes Paul
>> *Lähetetty:* keskiviikko 22. tammikuuta 2020 13.02
>> *Vastaanottaja:* mapserver-users at lists.osgeo.org
>> *Aihe:* [mapserver-users] cgi-error: End of script output before headers
>> mapserv
>>
>>
>>
>> Hello,
>>
>> my mapserver is having a very strange behaviour, which I can't explain,
>> nor debug.
>>
>> I'm using mapserver 7.0.0 along with apache http server 2.4.6.
>>
>>
>>
>> Most of the time, the server is working just fine, i.e. when I request a
>> typical URL, I get my image back:
>>
>>
>>
>>
>> http://10.0.0.22/cgi-bin/mapserv?map=/var/www/html/osm-transp-en-default.map&VERSION=1.1.0&REQUEST=getmap&SERVICE=wms&STYLES=&BBOX=-9.143,38.746,-9.140,38.748&HEIGHT=256&WIDTH=256&LAYERS=default&FORMAT=image/png&SRS=EPSG:4326&TRANSPARENT=true
>>
>>
>>
>> but then, when I slightly change bbox values, I get an internal server
>> error (500):
>>
>>
>>
>>
>> http://10.0.0.22/cgi-bin/mapserv?map=/var/www/html/osm-transp-en-default.map&VERSION=1.1.0&REQUEST=getmap&SERVICE=wms&STYLES=&BBOX=-9.1433,38.7460,-9.1406,38.7487&HEIGHT=256&WIDTH=256&LAYERS=default&FORMAT=image/png&SRS=EPSG:4326&TRANSPARENT=true
>>
>>
>>
>> I don't get any mapserver error log, but I do get an apache error log
>> saying:
>>
>> [cgi:error] End of script output before headers: mapserv
>>
>>
>>
>> This is really an strange behaviour which I can't explain.
>>
>> Did anyone already experienced a similar situation ?
>>
>> Thanks
>>
>> Jo
>> _______________________________________________
>> mapserver-users mailing list
>> mapserver-users at lists.osgeo.org
>> https://lists.osgeo.org/mailman/listinfo/mapserver-users
>
> _______________________________________________
> mapserver-users mailing list
> mapserver-users at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/mapserver-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/mapserver-users/attachments/20200123/3dfaf326/attachment.htm>