I have an interesting problem which I'm guessing is a bug but I'm not
sure where. One of my student's projects
(http://webmapping.mgis.psu.edu/geoserver/www/atg.html) used some
GeoTiffs this semester (he's the first to do this) and I upgraded to
GeoServer 1.7.1 at the same time. Now whenever I reboot the machine
(or the power goes out - which is not uncommon in rural PA) when
GeoServer comes back up it has lost the named coverage directory for
each of the geotiffs in the coverages directory but not the other
coverages. The other coverages were added using 1.6.5 (it's a shared
data directory).
As far as I can see the directories and info.xml files are fine when
the coverages are created. My only guess is that they are never
properly closed when they are created???
$ uname -a
Linux webmapping.ems.psu.edu 2.6.9-78.0.13.EL #1 Wed Jan 7 17:31:05
EST 2009 i686 i686 i386 GNU/Linux
Let me know if I should file a bug or if you'd like me to try any other tests?
Ian
On Thu, Jan 22, 2009 at 11:36 AM, Ian Turton <ijturton@anonymised.com> wrote:
I have an interesting problem which I'm guessing is a bug but I'm not
sure where. One of my student's projects
(http://webmapping.mgis.psu.edu/geoserver/www/atg.html) used some
GeoTiffs this semester (he's the first to do this) and I upgraded to
GeoServer 1.7.1 at the same time. Now whenever I reboot the machine
(or the power goes out - which is not uncommon in rural PA) when
GeoServer comes back up it has lost the named coverage directory for
each of the geotiffs in the coverages directory but not the other
coverages. The other coverages were added using 1.6.5 (it's a shared
data directory).
As far as I can see the directories and info.xml files are fine when
the coverages are created. My only guess is that they are never
properly closed when they are created???
Probably not, but just to double-check: these coverages don't exist in
the /tmp directory, do they?
Alex
On Thu, Jan 22, 2009 at 5:01 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
As far as I can see the directories and info.xml files are fine when
the coverages are created. My only guess is that they are never
properly closed when they are created???
Probably not, but just to double-check: these coverages don't exist in
the /tmp directory, do they?
Here are all the tmp directories I can find:
-bash-3.00$ ls /tmp
geowebcache.log hsperfdata_root logfile vmware-config0 vmware-root
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/
Databases
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/
HSQL
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/HSQL/
EPSG.data EPSG.properties EPSG.script
Ian
On 1/22/09, Ian Turton <ijturton@anonymised.com> wrote:
On Thu, Jan 22, 2009 at 5:01 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
>>
>> As far as I can see the directories and info.xml files are fine when
>> the coverages are created. My only guess is that they are never
>> properly closed when they are created???
>
> Probably not, but just to double-check: these coverages don't exist in
> the /tmp directory, do they?
Here are all the tmp directories I can find:
-bash-3.00$ ls /tmp
geowebcache.log hsperfdata_root logfile vmware-config0 vmware-root
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/
Databases
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/
HSQL
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/HSQL/
EPSG.data EPSG.properties EPSG.script
Ian
The reason I asked you this is because distros nowadays blow away
everything in /tmp on boot.
Alex
On Thu, Jan 22, 2009 at 8:40 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
On 1/22/09, Ian Turton <ijturton@anonymised.com> wrote:
On Thu, Jan 22, 2009 at 5:01 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
>>
>> As far as I can see the directories and info.xml files are fine when
>> the coverages are created. My only guess is that they are never
>> properly closed when they are created???
>
> Probably not, but just to double-check: these coverages don't exist in
> the /tmp directory, do they?
Here are all the tmp directories I can find:
-bash-3.00$ ls /tmp
geowebcache.log hsperfdata_root logfile vmware-config0 vmware-root
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/
Databases
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/
HSQL
-bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/HSQL/
EPSG.data EPSG.properties EPSG.script
Ian
The reason I asked you this is because distros nowadays blow away
everything in /tmp on boot.
Ah I see - No the coverages are built in the right directory - they
just vanish on reboot or I've discovered a tomcat restart.
Ian
On 1/22/09, Ian Turton <ijturton@anonymised.com> wrote:
On Thu, Jan 22, 2009 at 8:40 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
> On 1/22/09, Ian Turton <ijturton@anonymised.com> wrote:
>> On Thu, Jan 22, 2009 at 5:01 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
>>
>> >>
>> >> As far as I can see the directories and info.xml files are fine when
>> >> the coverages are created. My only guess is that they are never
>> >> properly closed when they are created???
>> >
>> > Probably not, but just to double-check: these coverages don't exist in
>> > the /tmp directory, do they?
>>
>>
>> Here are all the tmp directories I can find:
>>
>> -bash-3.00$ ls /tmp
>> geowebcache.log hsperfdata_root logfile vmware-config0 vmware-root
>> -bash-3.00$ ls /usr/local/tomcat/temp/Geotools/
>> Databases
>> -bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/
>> HSQL
>> -bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/HSQL/
>> EPSG.data EPSG.properties EPSG.script
>>
>>
>> Ian
>
> The reason I asked you this is because distros nowadays blow away
> everything in /tmp on boot.
>
Ah I see - No the coverages are built in the right directory - they
just vanish on reboot or I've discovered a tomcat restart.
Another thing: do the rasters reside somewhere in tomcat's
webapps/geoserver directory? It might be that if the geoserver.war
file is in the webapps directory, the uncompressed archive gets
refreshed from the war on tomcat startup.
Alex
On Thu, Jan 22, 2009 at 9:56 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
On 1/22/09, Ian Turton <ijturton@anonymised.com> wrote:
On Thu, Jan 22, 2009 at 8:40 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
> On 1/22/09, Ian Turton <ijturton@anonymised.com> wrote:
>> On Thu, Jan 22, 2009 at 5:01 PM, Alexander Petkov <greenkov@anonymised.com> wrote:
>>
>> >>
>> >> As far as I can see the directories and info.xml files are fine when
>> >> the coverages are created. My only guess is that they are never
>> >> properly closed when they are created???
>> >
>> > Probably not, but just to double-check: these coverages don't exist in
>> > the /tmp directory, do they?
>>
>>
>> Here are all the tmp directories I can find:
>>
>> -bash-3.00$ ls /tmp
>> geowebcache.log hsperfdata_root logfile vmware-config0 vmware-root
>> -bash-3.00$ ls /usr/local/tomcat/temp/Geotools/
>> Databases
>> -bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/
>> HSQL
>> -bash-3.00$ ls /usr/local/tomcat/temp/Geotools/Databases/HSQL/
>> EPSG.data EPSG.properties EPSG.script
>>
>>
>> Ian
>
> The reason I asked you this is because distros nowadays blow away
> everything in /tmp on boot.
>
Ah I see - No the coverages are built in the right directory - they
just vanish on reboot or I've discovered a tomcat restart.
Another thing: do the rasters reside somewhere in tomcat's
webapps/geoserver directory? It might be that if the geoserver.war
file is in the webapps directory, the uncompressed archive gets
refreshed from the war on tomcat startup.
No they live in webapps/geoserver-data as a shared data folder and
none of the other coverages or feature types are lost on restart just
the geotiffs that were added using 1.7.1
Ian