Hi,
In following patch I am using ctypes to load libgfapi(GlusterFS API) to get the Gluster volume statvfs information. http://gerrit.ovirt.org/#/c/17822
I am getting *segfault* when I run vdsClient 0 glusterVolumeSizeInfoGet volumeName=gv1
But I checked as below and it is working.
cd /usr/share/vdsm python >> from gluster import gfapi >> print gfapi.volumeStatvfs("gv1") posix.statvfs_result(f_bsize=4096L, f_frsize=4096L, f_blocks=25803070L, f_bfree=19130426L, f_bavail=17819706L, f_files=6553600L, f_ffree=5855876L, f_favail=5855876L, f_flag=4096L, f_namemax=255L)
Please suggest how I can debug this issue.
-- Regards Aravinda
I haven't worked much on using ctypes in python.. but did you try to look into the core to see where exactly the segfault is happening.. that might give some clues for further debug.
On 08/22/2013 05:53 PM, Aravinda wrote:
Hi,
In following patch I am using ctypes to load libgfapi(GlusterFS API) to get the Gluster volume statvfs information. http://gerrit.ovirt.org/#/c/17822
I am getting *segfault* when I run vdsClient 0 glusterVolumeSizeInfoGet volumeName=gv1
But I checked as below and it is working.
cd /usr/share/vdsm python >> from gluster import gfapi >> print gfapi.volumeStatvfs("gv1") posix.statvfs_result(f_bsize=4096L, f_frsize=4096L,
f_blocks=25803070L, f_bfree=19130426L, f_bavail=17819706L, f_files=6553600L, f_ffree=5855876L, f_favail=5855876L, f_flag=4096L, f_namemax=255L)
Please suggest how I can debug this issue.
-- Regards Aravinda
vdsm-devel mailing list vdsm-devel@lists.fedorahosted.org https://lists.fedorahosted.org/mailman/listinfo/vdsm-devel
Maybe you should pass it with quotes -> vdsClient 0 glusterVolumeSizeInfoGet volumeName="gv1" But probably that's not the issue, Did you try to add prints in http://gerrit.ovirt.org/#/c/17822/5/client/vdsClientGluster.py to see where the segfault is raised exactly ? still, if you prefer to debug step by step you can use any remote debuger (as pycharm) and in this case its easy, as it only one process and direct flow to step on. You need to check the all flow and not only the last call (gfapi.volumeStatvfs("gv1")) If you sure it raises inside the gfapi call with the same "gv1" as input, let me know and I'll investigate more
Yaniv Bronhaim
----- Original Message -----
From: "Aravinda" avishwan@redhat.com To: vdsm-devel@lists.fedorahosted.org Sent: Thursday, August 22, 2013 3:23:34 PM Subject: [vdsm] Help needed to debug segfault when using ctypes
Hi,
In following patch I am using ctypes to load libgfapi(GlusterFS API) to get the Gluster volume statvfs information. http://gerrit.ovirt.org/#/c/17822
I am getting segfault when I run vdsClient 0 glusterVolumeSizeInfoGet volumeName=gv1
But I checked as below and it is working.
cd /usr/share/vdsm python
from gluster import gfapi print gfapi.volumeStatvfs("gv1")
posix.statvfs_result(f_bsize=4096L, f_frsize=4096L, f_blocks=25803070L, f_bfree=19130426L, f_bavail=17819706L, f_files=6553600L, f_ffree=5855876L, f_favail=5855876L, f_flag=4096L, f_namemax=255L)
Please suggest how I can debug this issue.
-- Regards Aravinda
vdsm-devel mailing list vdsm-devel@lists.fedorahosted.org https://lists.fedorahosted.org/mailman/listinfo/vdsm-devel
On 08/27/2013 12:03 PM, Yaniv Bronheim wrote:
Maybe you should pass it with quotes -> vdsClient 0 glusterVolumeSizeInfoGet volumeName="gv1" But probably that's not the issue, Did you try to add prints in http://gerrit.ovirt.org/#/c/17822/5/client/vdsClientGluster.py to see where the segfault is raised exactly ?
No issue with vdsClient, I executed vdsm verb as below and got the same error. import xmlrpclib proxy = xmlrpclib.ServerProxy("http://localhost:54321") proxy.glusterVolumeSizeInfoGet("gv1")
still, if you prefer to debug step by step you can use any remote debuger (as pycharm) and in this case its easy, as it only one process and direct flow to step on. You need to check the all flow and not only the last call (gfapi.volumeStatvfs("gv1"))
gfapi.volumeStatvfs has all the code related to loading/using libgfapi. This func is added as supervdsm method using the decorator @makePublic, which is used in /vdsm/gluster/api.py (self.svdsmProxy.glusterVolumeStatvfs(volumeName))
Will try to debug using remote debugger. Thanks for the suggestions.
-- regards Aravinda
If you sure it raises inside the gfapi call with the same "gv1" as input, let me know and I'll investigate more
Yaniv Bronhaim
----- Original Message -----
From: "Aravinda" avishwan@redhat.com To: vdsm-devel@lists.fedorahosted.org Sent: Thursday, August 22, 2013 3:23:34 PM Subject: [vdsm] Help needed to debug segfault when using ctypes
Hi,
In following patch I am using ctypes to load libgfapi(GlusterFS API) to get the Gluster volume statvfs information. http://gerrit.ovirt.org/#/c/17822
I am getting segfault when I run vdsClient 0 glusterVolumeSizeInfoGet volumeName=gv1
But I checked as below and it is working.
cd /usr/share/vdsm python
from gluster import gfapi print gfapi.volumeStatvfs("gv1")
posix.statvfs_result(f_bsize=4096L, f_frsize=4096L, f_blocks=25803070L, f_bfree=19130426L, f_bavail=17819706L, f_files=6553600L, f_ffree=5855876L, f_favail=5855876L, f_flag=4096L, f_namemax=255L)
Please suggest how I can debug this issue.
-- Regards Aravinda
vdsm-devel mailing list vdsm-devel@lists.fedorahosted.org https://lists.fedorahosted.org/mailman/listinfo/vdsm-devel
Can you share vdsm.log and supervdsm.log , maybe I can find there something about the exception..
Yaniv Bronhaim.
----- Original Message -----
From: "Aravinda" avishwan@redhat.com To: "Yaniv Bronheim" ybronhei@redhat.com Cc: vdsm-devel@lists.fedorahosted.org Sent: Tuesday, August 27, 2013 12:29:40 PM Subject: Re: [vdsm] Help needed to debug segfault when using ctypes
On 08/27/2013 12:03 PM, Yaniv Bronheim wrote:
Maybe you should pass it with quotes -> vdsClient 0 glusterVolumeSizeInfoGet volumeName="gv1" But probably that's not the issue, Did you try to add prints in http://gerrit.ovirt.org/#/c/17822/5/client/vdsClientGluster.py to see where the segfault is raised exactly ?
No issue with vdsClient, I executed vdsm verb as below and got the same error. import xmlrpclib proxy = xmlrpclib.ServerProxy("http://localhost:54321") proxy.glusterVolumeSizeInfoGet("gv1")
still, if you prefer to debug step by step you can use any remote debuger (as pycharm) and in this case its easy, as it only one process and direct flow to step on. You need to check the all flow and not only the last call (gfapi.volumeStatvfs("gv1"))
gfapi.volumeStatvfs has all the code related to loading/using libgfapi. This func is added as supervdsm method using the decorator @makePublic, which is used in /vdsm/gluster/api.py (self.svdsmProxy.glusterVolumeStatvfs(volumeName))
Will try to debug using remote debugger. Thanks for the suggestions.
-- regards Aravinda
If you sure it raises inside the gfapi call with the same "gv1" as input, let me know and I'll investigate more
Yaniv Bronhaim
----- Original Message -----
From: "Aravinda" avishwan@redhat.com To: vdsm-devel@lists.fedorahosted.org Sent: Thursday, August 22, 2013 3:23:34 PM Subject: [vdsm] Help needed to debug segfault when using ctypes
Hi,
In following patch I am using ctypes to load libgfapi(GlusterFS API) to get the Gluster volume statvfs information. http://gerrit.ovirt.org/#/c/17822
I am getting segfault when I run vdsClient 0 glusterVolumeSizeInfoGet volumeName=gv1
But I checked as below and it is working.
cd /usr/share/vdsm python
from gluster import gfapi print gfapi.volumeStatvfs("gv1")
posix.statvfs_result(f_bsize=4096L, f_frsize=4096L, f_blocks=25803070L, f_bfree=19130426L, f_bavail=17819706L, f_files=6553600L, f_ffree=5855876L, f_favail=5855876L, f_flag=4096L, f_namemax=255L)
Please suggest how I can debug this issue.
-- Regards Aravinda
vdsm-devel mailing list vdsm-devel@lists.fedorahosted.org https://lists.fedorahosted.org/mailman/listinfo/vdsm-devel
vdsm-devel@lists.fedorahosted.org