[Geoserver-users] Geoserver crashing

Suvajit · July 1, 2009, 3:13pm

Hi
I am running Geoserver 1.7.3.a on Linux RHEL4. I am using JRE 1.6.
I am finding this problem recently where the JVM is crashing with messages
like as follows :

#
# An unexpected error has been detected by Java Runtime Environment:
#
# Internal Error (constantPoolOop.hpp:233), pid=24095, tid=3086879664
# Error: guarantee(tag_at(which).is_klass(),"Corrupted constant pool")
#
# Java VM: Java HotSpot(TM) Client VM (10.0-b23 mixed mode, sharing
linux-x86)
# An error report file with more information is saved as:
# /usr/local/geoserver/hs_err_pid24095.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp

My JVM arguments are as follows :

jvm_args: -DGEOSERVER_DATA_DIR=/usr/local/geoserver/data_dir -Xms64m
-Xmx256m -DSTOP.KEY=geoserver
java_command: start.jar
Launcher Type: SUN_STANDARD

As far I can decipher the error log, the problem seems to be that JVM is
using up the whole heap space.
See error log excerpt below:

Heap
def new generation total 4608K, used 4608K [0x802b0000, 0x807b0000,
0x81660000)
  eden space 4096K, 100% used [0x802b0000, 0x806b0000, 0x806b0000)
  from space 512K, 100% used [0x80730000, 0x807b0000, 0x807b0000)
  to space 512K, 100% used [0x806b0000, 0x80730000, 0x80730000)
tenured generation total 60544K, used 40742K [0x81660000, 0x85180000,
0x902b0000)
   the space 60544K, 67% used [0x81660000, 0x83e29b58, 0x83e29c00,
0x85180000)
compacting perm gen total 24320K, used 24122K [0x902b0000, 0x91a70000,
0x942b0000)
   the space 24320K, 99% used [0x902b0000, 0x91a3ea68, 0x91a3ec00,
0x91a70000)
    ro space 8192K, 69% used [0x942b0000, 0x94840780, 0x94840800,
0x94ab0000)
    rw space 12288K, 57% used [0x94ab0000, 0x951a0300, 0x951a0400,
0x956b0000)

Under this circumstances what can be done to save Geoserver from crashing ?

Regards,
Suvajit

--
View this message in context: http://www.nabble.com/Geoserver-crashing-tp24291585p24291585.html
Sent from the GeoServer - User mailing list archive at Nabble.com.

Suvajit · July 1, 2009, 3:18pm

I have also tried allocating 512m to JVM but Geoserver still crashes.
Only difference is that the gap between two consecutive crashes is widened

Suvajit wrote:

Hi
I am running Geoserver 1.7.3.a on Linux RHEL4. I am using JRE 1.6.
I am finding this problem recently where the JVM is crashing with messages
like as follows :

#
# An unexpected error has been detected by Java Runtime Environment:
#
# Internal Error (constantPoolOop.hpp:233), pid=24095, tid=3086879664
# Error: guarantee(tag_at(which).is_klass(),"Corrupted constant pool")
#
# Java VM: Java HotSpot(TM) Client VM (10.0-b23 mixed mode, sharing
linux-x86)
# An error report file with more information is saved as:
# /usr/local/geoserver/hs_err_pid24095.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp

My JVM arguments are as follows :

jvm_args: -DGEOSERVER_DATA_DIR=/usr/local/geoserver/data_dir -Xms64m
-Xmx256m -DSTOP.KEY=geoserver
java_command: start.jar
Launcher Type: SUN_STANDARD

As far I can decipher the error log, the problem seems to be that JVM is
using up the whole heap space.
See error log excerpt below:

Heap
def new generation total 4608K, used 4608K [0x802b0000, 0x807b0000,
0x81660000)
  eden space 4096K, 100% used [0x802b0000, 0x806b0000, 0x806b0000)
  from space 512K, 100% used [0x80730000, 0x807b0000, 0x807b0000)
  to space 512K, 100% used [0x806b0000, 0x80730000, 0x80730000)
tenured generation total 60544K, used 40742K [0x81660000, 0x85180000,
0x902b0000)
   the space 60544K, 67% used [0x81660000, 0x83e29b58, 0x83e29c00,
0x85180000)
compacting perm gen total 24320K, used 24122K [0x902b0000, 0x91a70000,
0x942b0000)
   the space 24320K, 99% used [0x902b0000, 0x91a3ea68, 0x91a3ec00,
0x91a70000)
    ro space 8192K, 69% used [0x942b0000, 0x94840780, 0x94840800,
0x94ab0000)
    rw space 12288K, 57% used [0x94ab0000, 0x951a0300, 0x951a0400,
0x956b0000)

Under this circumstances what can be done to save Geoserver from crashing
?

Regards,
Suvajit

--
View this message in context: http://www.nabble.com/Geoserver-crashing-tp24291585p24291677.html
Sent from the GeoServer - User mailing list archive at Nabble.com.

Justin_Deoliveira · July 2, 2009, 6:02am

You may find this recent read relevant:

http://www.nabble.com/JVM-1.6.0_06-crash%2C-libjvm.so%2B0x2641d8-td24177885.html#a24177885

-Justin

Suvajit wrote:

I have also tried allocating 512m to JVM but Geoserver still crashes.
Only difference is that the gap between two consecutive crashes is widened

Suvajit wrote:

Hi
I am running Geoserver 1.7.3.a on Linux RHEL4. I am using JRE 1.6.
I am finding this problem recently where the JVM is crashing with messages
like as follows :

#
# An unexpected error has been detected by Java Runtime Environment:
#
# Internal Error (constantPoolOop.hpp:233), pid=24095, tid=3086879664
# Error: guarantee(tag_at(which).is_klass(),"Corrupted constant pool")
#
# Java VM: Java HotSpot(TM) Client VM (10.0-b23 mixed mode, sharing
linux-x86)
# An error report file with more information is saved as:
# /usr/local/geoserver/hs_err_pid24095.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp

My JVM arguments are as follows :

jvm_args: -DGEOSERVER_DATA_DIR=/usr/local/geoserver/data_dir -Xms64m
-Xmx256m -DSTOP.KEY=geoserver
java_command: start.jar
Launcher Type: SUN_STANDARD

As far I can decipher the error log, the problem seems to be that JVM is
using up the whole heap space.
See error log excerpt below:

Heap
def new generation total 4608K, used 4608K [0x802b0000, 0x807b0000,
0x81660000)
  eden space 4096K, 100% used [0x802b0000, 0x806b0000, 0x806b0000)
  from space 512K, 100% used [0x80730000, 0x807b0000, 0x807b0000)
  to space 512K, 100% used [0x806b0000, 0x80730000, 0x80730000)
tenured generation total 60544K, used 40742K [0x81660000, 0x85180000,
0x902b0000)
   the space 60544K, 67% used [0x81660000, 0x83e29b58, 0x83e29c00,
0x85180000)
compacting perm gen total 24320K, used 24122K [0x902b0000, 0x91a70000,
0x942b0000)
   the space 24320K, 99% used [0x902b0000, 0x91a3ea68, 0x91a3ec00,
0x91a70000)
    ro space 8192K, 69% used [0x942b0000, 0x94840780, 0x94840800,
0x94ab0000)
    rw space 12288K, 57% used [0x94ab0000, 0x951a0300, 0x951a0400,
0x956b0000)

Under this circumstances what can be done to save Geoserver from crashing
?

Regards,
Suvajit

--
Justin Deoliveira
OpenGeo - http://opengeo.org
Enterprise support for open source geospatial.

Suvajit · July 2, 2009, 6:11am

Hi Justin,
  Thanks for the pointer.
  As said in
http://geoserver.org/display/GEOSDOC/2.6+GeoServer+in+Production+Environment
  I have tried to set off MALLOC_CHECK.
  But still Geoserver crashes.
Regards,
Suvajit

Justin Deoliveira-6 wrote:

You may find this recent read relevant:

http://www.nabble.com/JVM-1.6.0_06-crash%2C-libjvm.so%2B0x2641d8-td24177885.html#a24177885

-Justin

Suvajit wrote:

I have also tried allocating 512m to JVM but Geoserver still crashes.
Only difference is that the gap between two consecutive crashes is
widened

Suvajit wrote:

Hi
I am running Geoserver 1.7.3.a on Linux RHEL4. I am using JRE 1.6.
I am finding this problem recently where the JVM is crashing with
messages
like as follows :

#
# An unexpected error has been detected by Java Runtime Environment:
#
# Internal Error (constantPoolOop.hpp:233), pid=24095, tid=3086879664
# Error: guarantee(tag_at(which).is_klass(),"Corrupted constant pool")
#
# Java VM: Java HotSpot(TM) Client VM (10.0-b23 mixed mode, sharing
linux-x86)
# An error report file with more information is saved as:
# /usr/local/geoserver/hs_err_pid24095.log
#
# If you would like to submit a bug report, please visit:
# http://java.sun.com/webapps/bugreport/crash.jsp

My JVM arguments are as follows :

jvm_args: -DGEOSERVER_DATA_DIR=/usr/local/geoserver/data_dir -Xms64m
-Xmx256m -DSTOP.KEY=geoserver
java_command: start.jar
Launcher Type: SUN_STANDARD

As far I can decipher the error log, the problem seems to be that JVM is
using up the whole heap space.
See error log excerpt below:

Heap
def new generation total 4608K, used 4608K [0x802b0000, 0x807b0000,
0x81660000)
  eden space 4096K, 100% used [0x802b0000, 0x806b0000, 0x806b0000)
  from space 512K, 100% used [0x80730000, 0x807b0000, 0x807b0000)
  to space 512K, 100% used [0x806b0000, 0x80730000, 0x80730000)
tenured generation total 60544K, used 40742K [0x81660000, 0x85180000,
0x902b0000)
   the space 60544K, 67% used [0x81660000, 0x83e29b58, 0x83e29c00,
0x85180000)
compacting perm gen total 24320K, used 24122K [0x902b0000, 0x91a70000,
0x942b0000)
   the space 24320K, 99% used [0x902b0000, 0x91a3ea68, 0x91a3ec00,
0x91a70000)
    ro space 8192K, 69% used [0x942b0000, 0x94840780, 0x94840800,
0x94ab0000)
    rw space 12288K, 57% used [0x94ab0000, 0x951a0300, 0x951a0400,
0x956b0000)

Under this circumstances what can be done to save Geoserver from
crashing
?

Regards,
Suvajit

--
Justin Deoliveira
OpenGeo - http://opengeo.org
Enterprise support for open source geospatial.

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

--
View this message in context: http://www.nabble.com/Geoserver-crashing-tp24291585p24301725.html
Sent from the GeoServer - User mailing list archive at Nabble.com.

Andrea_Aime3 · July 2, 2009, 6:47am

Suvajit ha scritto:

Hi Justin,
  Thanks for the pointer. As said in
http://geoserver.org/display/GEOSDOC/2.6+GeoServer+in+Production+Environment
  I have tried to set off MALLOC_CHECK.
  But still Geoserver crashes.
Regards,
Suvajit

Are you using any native extension, such as the GDAL one to read ECW or
MrSID imagery?
If not the problem is certainly the VM fault itself, java is not like
C, there is nothing you can do past accessing native code directly
that can take down the VM unless the VM itself is buggy or, as in the
case of the malloc check, the base library suddenly decide to change
behavior breaking the upper levels.

Did you try upgrading the virtual machine? Oh, btw, to run GeoServer
1.7.x you need a JDK, not a JRE.

Another thing missing there is the failure trace, the part that's
labelled "# Problematic frame". Would be interesting to see that as
well.

Cheers
Andrea

--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.

Suvajit_Sengupta · July 2, 2009, 6:56am

Hi Andrea,

No I am not using any GDAL extensions. There is no raster in my dataset.
I will switch to JDK.
Here are some problematic frames of consecutive crashes:

Problematic frame:

V [libjvm.so+0x35afa8]

Problematic frame:

V [libjvm.so+0x1b9011]

Problematic frame:

V [libjvm.so+0x2ee27d]

Regards,
Suvajit

My galleria @ http://www.flickr.com/photos/jeet_sen

On Thu, Jul 2, 2009 at 12:17 PM, Andrea Aime <aaime@anonymised.com1…> wrote:

Suvajit ha scritto:

Hi Justin,
Thanks for the pointer. As said in
http://geoserver.org/display/GEOSDOC/2.6+GeoServer+in+Production+Environment
I have tried to set off MALLOC_CHECK.
But still Geoserver crashes.
Regards,
Suvajit

Are you using any native extension, such as the GDAL one to read ECW or
MrSID imagery?
If not the problem is certainly the VM fault itself, java is not like
C, there is nothing you can do past accessing native code directly
that can take down the VM unless the VM itself is buggy or, as in the
case of the malloc check, the base library suddenly decide to change
behavior breaking the upper levels.

Did you try upgrading the virtual machine? Oh, btw, to run GeoServer
1.7.x you need a JDK, not a JRE.

Another thing missing there is the failure trace, the part that’s
labelled “# Problematic frame”. Would be interesting to see that as
well.

Cheers
Andrea

–
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.