Fwd: [MAPSERVER-USERS] Fwd: CLOSE_CONNECTION=DEFER - Segmentation fault

[Paul Ramsey]

Great work Dave. I will apply this pretty soon, unless I hear from you
that you've had negative reports from others.

P

On Fri, Apr 25, 2008 at 10:36 AM, Ivan Mincik <ivan.mincik at gmail.com> wrote:
> Thanks a lot Dave,
>  according to tests on my local machine there are no other problems with v6
>  patch. Next week I will try it on production system.
>
>  Ivan
>
>  On Sunday 20 April 2008 21:26, you wrote:
>  > Ivan,
>  >
>  >    BTW there is a new version of the mappostgis.c patch, v6 at
>  > http://trac.osgeo.org/mapserver/ticket/2497 , which further cleans up
>  > the code (but doesn't add any new enhancements).  If you get a chance
>  > to test it and find any problems, let me know (and CC the
>  > mapserver-users list on reply).
>  >
>  > Thanks,
>  >
>  > Dave
>
>
> >
>  > On Sat, Mar 22, 2008 at 10:33 AM, Ivan Mincik <ivan.mincik at gmail.com> wrote:
>  > > Hi Dave and others,
>  > >
>  > >  >Here is the report about testing second version of
>  > >  >ms_postgis_begin_to_connect.diff (03/20/08 18:35:00) in p.mapper with
>  > >  >PROCESSING "CLOSE_CONNECTION=DEFER.
>  > >  >
>  > >  >I have started browsing the map, using Obj_queryBy functions, turning
>  > >  > on and off different layers. Application was working smoothly without
>  > >  > any error or warning in the log.
>  > >  >
>  > >  >THE PROBLEM will occur when I stop working in pmapper for a while and
>  > >  > try to continue browsing. (in the log it is between 09:41:44 and -
>  > >  > 09:46:05 and 09:50:03 - 09:52:30) . After these pauses I had to
>  > >  > restart browser to get p.mapper working again, but in 20 percent still
>  > >  > application is still freezing.
>  > >  >Same problems can be reproduced every new pmapper session.
>  > >
>  > >  I have solved this problem with php session settings so this last
>  > > problem was probably on php/p.mapper site. Now the logs are clear.
>  > > Tomorrow I will test it with more users then one.
>  > >  Ivan
>  > >
>  > >  >Maybe I can try some session configurations in apache or postgres. Any
>  > >  >advises?
>  > >  >
>  > >  >PostgreSQL log is here:
>  > >  >http://projects.gpsmapy.sk/logs/ms_postgis_begin_to_connect_version2.lo
>  > >  >g
>  > >  >
>  > >  >
>  > >  >Dave, thanks a lot for Your work
>  > >  >Ivan
>  > >  >
>  > >  >On Friday 21 March 2008 18:12, you wrote:
>  > >  > Ivan,
>  > >  >
>  > >  >    Yes, I updated the patch on the bug report page
>  > >  > http://trac.osgeo.org/mapserver/ticket/2497 around the time I sent
>  > >  > those emails yesterday.  Download it, make sure it differs from the
>  > >  > ms_postgis_fix_trans_state.diff you're currently using, `svn revert
>  > >  > mappostgis.c`, and try it with the new patch.
>  > >  >
>  > >  >    Sorry that was not clear.
>  > >  >
>  > >  > Thanks,
>  > >  >
>  > >  > Dave
>  > >  >
>  > >  > On Fri, Mar 21, 2008 at 4:59 AM, Ivan Mincik <ivan.mincik at gmail.com>
>  wrote:
>  > >  > > On Thursday 20 March 2008 23:46, you wrote:
>  > >  > >  > Ivan,
>  > >  > >  >
>  > >  > >  >    Try the new version of the patch, just posted.  I had
>  > >  > >  > accidentally removed a necessary query in msPOSTGISLayerGetShape.
>  > >  > >  >  My tests which now include getShape() are working (with or
>  > >  > >  > without PROCESSING "CLOSE_CONNECTION=DEFER").
>  > >  > >
>  > >  > >  Hi Dave,
>  > >  > >  I have allready tried patch ms_postgis_fix_trans_state.diff and I
>  > >  > > have posted the report. Or is there any updated version of
>  > >  > >  ms_postgis_fix_trans_state.diff?
>  > >  > >
>  > >  > >  I am sure that I have updated php_mapscript.so.
>  > >  > >
>  > >  > >  ************************************************************
>  > >  > >  For sure I include my report of testing
>  > >  > > ms_postgis_fix_trans_state.diff:
>  > >  > >
>  > >  > >
>  > >  > >  I have applied new patch ms_postgis_fix_trans_state.diff to
>  > >  > > mapserver-5.0.2 sources from debian unstable and recompiled new
>  > >  > > debian package and install.
>  > >  > >
>  > >  > >
>  > >  > >  Then I have:
>  > >  > >  1. restarted apache
>  > >  > >  2. restarted postgresql-8.1
>  > >  > >  3. restarted browser (deleted cache)
>  > >  > >
>  > >  > >  And started this test with p.mapper 3.2 application:
>  > >  > >
>  > >  > >  1. Browsing the map (moving, zooming out, zooming in, adding some
>  > >  > > more layers) 2. Using identify
>  > >  > > ({map,layer}Obj_queryBy{Point,Rect,etc.}) - ERROR: cursor "mycursor2
>  > >  > > immediately
>  > >  > >  3. Browsing the map (moving, zooming out, zooming in)
>  > >  > >  4. Using identify ({map,layer}Obj_queryBy{Point,Rect,etc.}) -
>  > >  > > ERROR: cursor "mycursor2 immediately
>  > >  > >
>  > >  > >
>  > >  > >  Result:
>  > >  > >  With this patch it is not possible to do neither a singel object
>  > >  > > query. (with older patch it was possible few times).
>  > >  > >  After "ERROR:  cursor "mycursor2" does not exist" it is possible to
>  > >  > > continue browsing the map again and object query fill finish with
>  > >  > > same result as above.
>  > >  > >
>  > >  > >
>  > >  > > {map,layer}Obj_queryBy{Point,Rect,etc.} function is not
>  > >  > >  working   also when  #PROCESSING "CLOSE_CONNECTION=DEFER" is
>  > >  > > disabled. So identify object is not working at all.
>  > >  > >
>  > >  > >  log is here:
>  > >  > >
>  > >  > > http://projects.gpsmapy.sk/logs/ms_postgis_fix_trans_state_patch.log
>  > >
>  > >  -------------------------------------------------------
>  _______________________________________________
>  mapserver-users mailing list
>  mapserver-users at lists.osgeo.org
>  http://lists.osgeo.org/mailman/listinfo/mapserver-users
>