Scott,
It looks like client-core sefaulted...
Lets do this then..
fire up pvfs2-client-core directly  in gdb after loading the module
and then issue the mount command. You should see exactly where it
crashes and why...

if you want logging output in the log file, do
pvfs2-client-core --gossip-mask="bmi,mx,..."
I am assuming that you have a keyword called mx that triggers
GOSSIP_BMI_MX_DEBUG or something?
I take it that servers are running and pvfs2-ping etc all work
flawlessly with the same setup..?
thanks,
Murali

On 1/2/07, Scott Atchley <[EMAIL PROTECTED]> wrote:
Hi Murali,

Yes, I am running the client and client-core daemons.

The client log is not very helpful:

% sudo cat /tmp/pvfs2-client.log
[E 11:41:22.939114] pvfs2-client-core with pid 14154 exited with value 1
[E 11:50:17.484256] pvfs2-client-core with pid 14241 exited with value 1
[E 16:08:25.934435] pvfs2-client-core with pid 15420 exited with value 1
[E 16:12:47.768309] pvfs2-client-core with pid 15450 exited with value 1

Dmesg shows a little more:

Jan  2 16:09:34 fog34 kernel: pvfs2-client-core: opening device
Jan  2 16:09:34 fog34 kernel: pvfs2-client-core: open device complete
(ret = 0)
Jan  2 16:09:34 fog34 kernel: pvfs2_bufmap_initialize: called (ptr
(b69c5000) sz (4194304) cnt(5).
Jan  2 16:09:34 fog34 kernel: pvfs2_bufmap_initialize: exiting normally
Jan  2 16:09:34 fog34 kernel: pvfs2_devreq_ioctl: got
PVFS_DEV_REMOUNT_ALL
Jan  2 16:09:34 fog34 kernel: pvfs2_devreq_ioctl: priority remount in
progress
Jan  2 16:09:34 fog34 kernel: pvfs2_devreq_ioctl: priority remount
complete

<at this point, I call mmount>

Jan  2 16:12:19 fog34 syslog-ng[2928]: STATS: dropped 0
Jan  2 16:12:47 fog34 kernel: pvfs2_get_sb: called with devname mx://
fog33:0:3/pvfs2-fs
Jan  2 16:12:47 fog34 kernel: Alloced OP (eb65a0a0: 101 OP_FS_MOUNT)
Jan  2 16:12:47 fog34 kernel: Attempting PVFS2 Mount via host mx://
fog33:0:3/pvfs2-fs
Jan  2 16:12:47 fog34 kernel: pvfs2: service_operation: pvfs2_get_sb
eb65a0a0
Jan  2 16:12:47 fog34 kernel: client-core: reading op tag 101
OP_FS_MOUNT
Jan  2 16:12:47 fog34 kernel: pvfs2-client-core: exiting, closing device
Jan  2 16:12:47 fog34 kernel: pvfs2_bufmap_finalize: called
Jan  2 16:12:47 fog34 kernel: pvfs2_bufmap_finalize: exiting normally
Jan  2 16:12:47 fog34 kernel: PVFS2 Device Close: Filesystem(s)
UNMOUNTED
Jan  2 16:12:47 fog34 kernel: pvfs2-client-core: purging in-progress
op tag 101 OP_FS_MOUNT
Jan  2 16:12:47 fog34 kernel: pvfs2-client-core: device close complete
Jan  2 16:12:47 fog34 kernel: Interrupted: Removed op eb65a0a0 from
htable_ops_in_progress
Jan  2 16:12:47 fog34 kernel: pvfs2: tag 101 (pvfs2_get_sb) --
operation to be retried (2 attempt)
Jan  2 16:12:47 fog34 kernel: pvfs2: service_operation: pvfs2_get_sb
eb65a0a0
Jan  2 16:13:07 fog34 kernel: *** operation timed out (tag 101,
eb65a0a0, att 2)
Jan  2 16:13:07 fog34 kernel: Interrupted: Removed op eb65a0a0 from
request_list
Jan  2 16:13:07 fog34 kernel: pvfs2: pvfs2_get_sb -- wait timed out;
aborting attempt.
Jan  2 16:13:07 fog34 kernel: pvfs2: service_operation pvfs2_get_sb
returning: -110 for eb65a0a0.
Jan  2 16:13:07 fog34 kernel: pvfs2_get_sb: mount got return value of
-110
Jan  2 16:13:07 fog34 kernel: pvfs2_get_sb: mount request failed with
-110
Jan  2 16:13:07 fog34 kernel: Releasing OP (eb65a0a0: 101)
Jan  2 16:13:07 fog34 kernel: pvfs2_get_sb: returning sb ffffff92

BMI_meth_method_addr_lookup() (nor any other BMI_meth* function) is
never called.

Scott

On Jan 2, 2007, at 3:33 PM, Murali Vilayannur wrote:

> Hi Scott,
> Are you running the client-core daemon which services the kmod
> requests in user space?..
> if not you  can run it like so
> pvfs2-client -p <path to pvfs2-client-core>
> All your fprints etc will be logged in /tmp/pvfs2-client.log if you
> enable the gossip mx debug mask when starting up client-core..
> hope this helps,
> Murali
>
> On 1/2/07, Scott Atchley <[EMAIL PROTECTED]> wrote:
>> Hi all,
>>
>> I want to run some of the test code. I have built the client kernel
>> module, installed the module and ran depmod. I can load it using
>> insmod or modprobe without any problems (shows up in lsmod output).
>>
>> When I try to mount the file system, however, I see:
>>
>> % sudo mount -t pvfs2 mx://fog33:0:3/pvfs2-fs /mnt/pvfs2
>> mount: Connection timed out
>>
>> I have fprintf()s all over my code and they are not being called.
>> Nothing shows on stdout or stderr in the terminal, but I do see this
>> in the client's dmesg:
>>
>> pvfs2: module version 2.6.1pre1-2007-01-02-162515 loaded
>> pvfs2: pvfs2_get_sb -- wait timed out; aborting attempt.
>> pvfs2_get_sb: mount request failed with -110
>>
>> Error -110 is:
>>
>> #define ETIMEDOUT       110     /* Connection timed out */
>>
>> I then loaded pvfs2.ko with gossip_debug=-1. I still see the timeout
>> and dmesg now has:
>>
>> pvfs2: pvfs2_init called with debug mask 0xffffffff
>> *** /dev/pvfs2-req character device registered ***
>> 'mknod /dev/pvfs2-req c 254 0'.
>> pvfs2: module version 2.6.1pre1-2007-01-02-162515 loaded
>> pvfs2_get_sb: called with devname mx://fog33:0:3/pvfs2-fs
>> Alloced OP (f3cf4020: 100 OP_FS_MOUNT)
>> Attempting PVFS2 Mount via host mx://fog33:0:3/pvfs2-fs
>> pvfs2: service_operation: pvfs2_get_sb f3cf4020
>> *** operation timed out (tag 100, f3cf4020, att 1)
>> Interrupted: Removed op f3cf4020 from request_list
>> pvfs2: pvfs2_get_sb -- wait timed out; aborting attempt.
>> pvfs2: service_operation pvfs2_get_sb returning: -110 for f3cf4020.
>> pvfs2_get_sb: mount got return value of -110
>> pvfs2_get_sb: mount request failed with -110
>> Releasing OP (f3cf4020: 100)
>> pvfs2_get_sb: returning sb ffffff92
>>
>> What am I doing wrong?
>>
>> Thanks,
>>
>> Scott
>> _______________________________________________
>> Pvfs2-developers mailing list
>> [email protected]
>> http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers
>>


_______________________________________________
Pvfs2-developers mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers

Reply via email to