[Geoserver-users] Geoserver 2.1.3 2.2 some problems and questions

Hello there,

I will be very happy if you can answer these questions and problems.
We have been using Geoserver 2.1.3 and then 2.2 but we faced with some
problems while using it in our projects. They are:

1- What if table is empty? There will be no SRID defined (in mssql 2008 r2)
so that might cause problems? Is it safe to add empty tables to geoserver?

1.2- What if there are some rows in the table which have null geometry? That
causes a problem?

2- What if there are geometries which are not valid? Is it safe or not safe?

3- What if there are some geometry collections in a table? What is the
impact? As far as I know shape files cannot accept "geometry collection" but
is it harmful for geoserver as well?

4- We have spatial indexes in all tables in MSSQL 2008 r2 but we are getting
data from views of these tables, still spatial indexes will work or not?

5- Let's say 100 row of table has SRID 4326 but one row has 900913 then what
is the impact on geoserver if we add this table.

6- We have some tables that geoserver cannot use its spatial index because
of a bug in database server. Mssql 2008 r2 bug happens when there are many
cores in the machine (128 in our case), so query optimizer cannot understand
spatial index usage will be helpful then tries to use normal index. Is there
a way to force geoserver to use particular indexes in mssql? For example :
with(index(x index)) command etc.

7- What can be the cause of "PermGen space" error? There is enough memory
for permgen but periodically we are getting this error. That can be because
of some problematic data? That happens mostly in Geoserver 2.2 and I
couldn't find bug reports related with it.

I am asking these questions because either 2.1.3 or 2.2 every few hours
geoserver becomes unreachable. Thank you already,

Best Regards.

--
View this message in context: http://osgeo-org.1560.n6.nabble.com/Geoserver-2-1-3-2-2-some-problems-and-questions-tp5020078.html
Sent from the GeoServer - User mailing list archive at Nabble.com.

Hello,

after this Email is lying around for a few days like a hot potato, I'll try
to answer a few questions. I do not have any experience with SQL Server I
might add, however, somebody else may be able to fill this in. What I can
offer is some knowledge, some experience and some opinion.

1 It is safe to add tables with no spatial column to geoserver. Yes. Things
like reference tables for instance, but you need to allow for this when
defining the datastore (tick the last tickbox 'geometryless tables'). With
the other question 'empty tables' uuuhm why would you do that?

2 I do not know with SQL Server, but not with other DBs.

3 I do not know but it seems to me a bad idea. How do you want to style
something like this?

4 Yes. At least they do with Oracle and ArcSDE

5 No geoserver will throw an error.

6 I do not know

7 java -server allows for a 64M permanent memory space. Geoserver has a lot
of classes. It needs more, particularly if you have many layers and styles
and plug-ins. So, java -XX:PermSize=96m -XX:MaxPermSize=128M. Something like
that, there are other mailings on this list dealing with this problem.

I do not know whether you have to use SQL Server or not, but it sounds to me
like a database system I would not touch. It appears to be a very lenient
system according to your description and one that allows bad geometry,
different geometries etc. One advice: please, please clean up your data, no
spikes, bowties, self-intersecting polygons etc. and any WMS will reward you
with a better stability. Geoserver is rather patient with these things but
it does not cater for everything.

You would want a WMS to be fast, but it will either be fast or do a good
exception handling. Esri is relatively stable and fast, but it is only so,
because ArcSDE is so finicky and fussy with valid geometry. You wouldn't
even get your data into ArcSDE. MapInfo has never managed to produce a
stable WMS. That is not because they cannot program, but because MapInfo
format accepts any spatial feature, however crooked.

Geoserver tries to occupy some middle ground as it caters for a range of
spatial formats. Please care about data quality and educate your data
providers otherwise operating a WMS, will be a nightmare whatever you use.

Cheers

Christian

-----
____________________________

Dr Christian Maul
Project Manager

Information Services Branch
Department of Sustainability and Environment
Level13, Marland House, 570 Bourke Street
Melbourne 3000

PO Box 500, East Melbourne Vic 3002

Telephone: +61-3-8636 2325
Telefax: +61-3-8636 2813
--
View this message in context: http://osgeo-org.1560.n6.nabble.com/Geoserver-2-1-3-2-2-some-problems-and-questions-tp5020078p5020809.html
Sent from the GeoServer - User mailing list archive at Nabble.com.

1- For empty tables, specific to MSSql server empty tables will give
exception like "SRID must be between 0 to 99999" which will break GeoServer.
This is serious problem since table can be empty sometimes (because of
pulling some active records to passive state).

--
View this message in context: http://osgeo-org.1560.n6.nabble.com/Geoserver-2-1-3-2-2-some-problems-and-questions-tp5020078p5028040.html
Sent from the GeoServer - User mailing list archive at Nabble.com.

I don’t know much about MSSQL, but if it breaks GeoServer this may be a bug.

Might be worth reporting it: http://jira.codehaus.org/browse/GEOS

On 17 January 2013 07:58, Sergeant_york <electronicpanda@anonymised.com> wrote:

1- For empty tables, specific to MSSql server empty tables will give
exception like “SRID must be between 0 to 99999” which will break GeoServer.
This is serious problem since table can be empty sometimes (because of
pulling some active records to passive state).


View this message in context: http://osgeo-org.1560.n6.nabble.com/Geoserver-2-1-3-2-2-some-problems-and-questions-tp5020078p5028040.html
Sent from the GeoServer - User mailing list archive at Nabble.com.


Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
MVPs and experts. ON SALE this month only – learn more at:
http://p.sf.net/sfu/learnmore_122712


Geoserver-users mailing list
Geoserver-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-users

This transmission is intended for the named addressee(s) only and may contain sensitive or protectively marked material up to RESTRICTED and should be handled accordingly. Unless you are the named addressee (or authorised to receive it for the addressee) you may not copy or use it, or disclose it to anyone else. If you have received this transmission in error please notify the sender immediately. All email traffic sent to or from us, including without limitation all GCSX traffic, may be subject to recording and/or monitoring in accordance with relevant legislation.

On Fri, Jan 18, 2013 at 11:49 AM, Jonathan Moules <jonathanmoules@…4942…> wrote:

I don’t know much about MSSQL, but if it breaks GeoServer this may be a bug.

Might be worth reporting it: http://jira.codehaus.org/browse/GEOS

Eh, the bug in is SQL Server, there is nothing we can do GeoServer side (afaik)

Cheers
Andrea

==
Our support, Your Success! Visit http://opensdi.geo-solutions.it for more information.

Ing. Andrea Aime
@geowolf
Technical Lead

GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 1660272
mob: +39 339 8844549

http://www.geo-solutions.it
http://twitter.com/geosolutions_it


On Thu, Jan 17, 2013 at 8:58 AM, Sergeant_york <electronicpanda@anonymised.com84…> wrote:

1- For empty tables, specific to MSSql server empty tables will give
exception like “SRID must be between 0 to 99999” which will break GeoServer.
This is serious problem since table can be empty sometimes (because of
pulling some active records to passive state).

SQL Server does not have a notion of “consistent srid” per table, which is a bit ridiculous.
To avoid this issue GeoServer can use a “geometry_columns” table where you manually
register the basic information about your table, including the srid and geometry type,
so that these information may be known also over an empty table (when there is at
least one record and no geometry_columns table GeoServer gets the first record
with a non null geometry and assumes all others are similar, which is kinda lame but
it’s as much as we can do with what SQL Server offers).

Unfortunately the above feature (the geometry metadata table) is only part of “trunk”,
what will become 2.3.0 in a couple of months

Cheers
Andrea

==
Our support, Your Success! Visit http://opensdi.geo-solutions.it for more information.

Ing. Andrea Aime
@geowolf
Technical Lead

GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 1660272
mob: +39 339 8844549

http://www.geo-solutions.it
http://twitter.com/geosolutions_it


Based on Andrea’s comment, I guess you’ll want to report it to your SQL Server support instead then.

On 18 January 2013 11:07, Andrea Aime <andrea.aime@anonymised.com> wrote:

On Fri, Jan 18, 2013 at 11:49 AM, Jonathan Moules <jonathanmoules@…4942…> wrote:

I don’t know much about MSSQL, but if it breaks GeoServer this may be a bug.

Might be worth reporting it: http://jira.codehaus.org/browse/GEOS

Eh, the bug in is SQL Server, there is nothing we can do GeoServer side (afaik)

Cheers
Andrea

==
Our support, Your Success! Visit http://opensdi.geo-solutions.it for more information.

Ing. Andrea Aime
@geowolf
Technical Lead

GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 1660272
mob: +39 339 8844549

http://www.geo-solutions.it
http://twitter.com/geosolutions_it


This transmission is intended for the named addressee(s) only and may contain sensitive or protectively marked material up to RESTRICTED and should be handled accordingly. Unless you are the named addressee (or authorised to receive it for the addressee) you may not copy or use it, or disclose it to anyone else. If you have received this transmission in error please notify the sender immediately. All email traffic sent to or from us, including without limitation all GCSX traffic, may be subject to recording and/or monitoring in accordance with relevant legislation.

On Sat, Dec 1, 2012 at 4:23 PM, Sergeant_york <electronicpanda@anonymised.com4…> wrote:

6- We have some tables that geoserver cannot use its spatial index because
of a bug in database server. Mssql 2008 r2 bug happens when there are many
cores in the machine (128 in our case), so query optimizer cannot understand
spatial index usage will be helpful then tries to use normal index. Is there
a way to force geoserver to use particular indexes in mssql? For example :
with(index(x index)) command etc.

There is no way to force GeoServer to use the index right now.
Some other user said that to force the usage of the index the SQL query should
be modified to contain an explicit query hint, this kind of modification would require
a bit of changes in the overall JDBC modules (most the code to access a spatial
database is shared between all databases, only differences are factored out
in specific classes, and query hint support is not there in the architecture to start with).

Not a massive undertaking by any means, but something that would require some
work for sure. Patches welcomed :slight_smile:

Cheers
Andrea

==
Our support, Your Success! Visit http://opensdi.geo-solutions.it for more information.

Ing. Andrea Aime
@geowolf
Technical Lead

GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 1660272
mob: +39 339 8844549

http://www.geo-solutions.it
http://twitter.com/geosolutions_it


Thank you for the answers,

For first situation, we solved the problem by inserting dummy geometries to
tables which will be visible after sld apply that is at left bottom corner
of our extent which is also kind of lame but worked well =)

Secondly, I have been searching is there an option for meta tables then I
got my answer, so dummy geometry solution will persist for a while it seems.

About hinting db for using spatial index; it is not so serious issue for any
other company which is running their own databases because microsoft
published a patch to fix the bug already. On the other hand, we are using a
database as an external source and sadly we cannot force db owner to apply
the patch.

Right now, for index issue, the solution seems using gwc as much as we can
to bypass speed issue. Unfortunately for the tables which will be changed
often, we have to find another solution.

Best regards,

--
View this message in context: http://osgeo-org.1560.n6.nabble.com/Geoserver-2-1-3-2-2-some-problems-and-questions-tp5020078p5028485.html
Sent from the GeoServer - User mailing list archive at Nabble.com.

On Sat, Jan 19, 2013 at 9:21 PM, Sergeant_york <electronicpanda@anonymised.com84…> wrote:

About hinting db for using spatial index; it is not so serious issue for any
other company which is running their own databases because microsoft
published a patch to fix the bug already. On the other hand, we are using a
database as an external source and sadly we cannot force db owner to apply
the patch.

Believe I’ve found the report you’re talking about here:
https://connect.microsoft.com/SQLServer/feedback/details/625896/spatial-index-not-used-on-multiproc-machine-unless-maxdop-set-downward

It’s marked as fixed, so I guess a hotfix package is already out for it, but
I could not locate it.
Pity, would have been good reference for SQL server users hit by this problem.

Cheers
Andrea

==
Our support, Your Success! Visit http://opensdi.geo-solutions.it for more information.

Ing. Andrea Aime
@geowolf
Technical Lead

GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584 962313
fax: +39 0584 1660272
mob: +39 339 8844549

http://www.geo-solutions.it
http://twitter.com/geosolutions_it