[Linux-cluster] Problems after upgrade :(

ptr at poczta.fm ptr at poczta.fm
Mon Apr 18 13:11:10 UTC 2005


    Hello.

I installed the newest CVS version from the scratch
(because of noticed problems with old libs and binaries
remaining in system directories even after "make install" 
with certain components. 
Anyways, now I'm getting those entries as below after node startup
and even during normal work:

GFS: Trying to join cluster "lock_dlm", "cluster1:eva"
scheduling while atomic: cman_comms/0x00000001/8808
 [<c039a4d2>] schedule+0xbc2/0xbd0
 [<c01156fe>] __wake_up+0x3e/0x60
 [<c039ba0f>] _spin_unlock_irqrestore+0xf/0x30
 [<f8bb5759>] queue_message+0x109/0x120 [cman]
 [<f8bbe06d>] add_barrier_callback+0x7d/0x160 [cman]
 [<f8bbe1a0>] callback_startdone_barrier_new+0x20/0x30 [cman]
 [<f8bb6d27>] check_barrier_complete_phase2+0xc7/0x110 [cman]
 [<f8bb6e45>] process_barrier_msg+0xa5/0x120 [cman]
 [<f8bb306f>] process_incoming_packet+0x18f/0x290 [cman]
 [<f8bb21b1>] receive_message+0xd1/0xf0 [cman]
 [<f8bb235c>] cluster_kthread+0x18c/0x340 [cman]
 [<c0115630>] default_wake_function+0x0/0x20
 [<f8bb21d0>] cluster_kthread+0x0/0x340 [cman]
 [<c01009f5>] kernel_thread_helper+0x5/0x10
scheduling while atomic: cman_comms/0x00000001/8808
 [<c039a4d2>] schedule+0xbc2/0xbd0
 [<f8bb26de>] start_ack_timer+0x2e/0x40 [cman]
 [<f8bbe06d>] add_barrier_callback+0x7d/0x160 [cman]
 [<f8bbe1a0>] callback_startdone_barrier_new+0x20/0x30 [cman]
 [<f8bb6d27>] check_barrier_complete_phase2+0xc7/0x110 [cman]
 [<f8bb6e45>] process_barrier_msg+0xa5/0x120 [cman]
 [<f8bb306f>] process_incoming_packet+0x18f/0x290 [cman]
 [<f8bb21b1>] receive_message+0xd1/0xf0 [cman]
 [<f8bb235c>] cluster_kthread+0x18c/0x340 [cman]
 [<c0115630>] default_wake_function+0x0/0x20
 [<f8bb21d0>] cluster_kthread+0x0/0x340 [cman]
 [<c01009f5>] kernel_thread_helper+0x5/0x10

Besides, sometimes when I reboot one of the nodes
(it's 2-nodes cluster running 2.6.11.7), it won't start up
showing on console messages like "CMANsendmsg failed: "-101".
I have to reboot again to start the node fully up.
    Any hints on what's wrong?

TIA for your help, best regards

Piotr




------------------------------------------------------------------
Teraz na tapecie mamy najwiekszego z silaczy. 
Sciagnij >> http://link.interia.pl/f1873 <<




More information about the Linux-cluster mailing list