OS : SUSE Linux Enterprise Server 11 (x86_64)
VERSION = 11
PATCHLEVEL = 2
Kernel : 3.0.13-0.27-default
oracleasm SUSE Linux Enterprise 11 2.0.5-7.24.59
oracleasm-kmp-default SUSE Linux Enterprise 11 2.0.5_3.0.13_0.27-7.24.59
kernel: [93272.425238] slab error in kmem_cache_destroy(): cache `asmdisk_cache': Can't free all objects
kernel: [93272.425248] Pid: 88504, comm: modprobe Tainted: G X 3.0.13-0.27-default #1
kernel: [93272.425253] Call Trace:
kernel: [93272.425314] [<ffffffff810048b5>] dump_trace+0x75/0x300
kernel: [93272.425334] [<ffffffff8143ea0f>] dump_stack+0x69/0x6f
kernel: [93272.425565] [<ffffffff8113db88>] kmem_cache_destroy+0xd8/0x120
kernel: [93272.425595] [<ffffffffa03439ca>] exit_asmfs_fs+0x1a/0x650 [oracleasm]
kernel: [93272.425649] [<ffffffff810975ee>] sys_delete_module+0x17e/0x270
kernel: [93272.425670] [<ffffffff81449692>] system_call_fastpath+0x16/0x1b
kernel: [93272.427580] DWARF2 unwinder stuck at system_call_fastpath+0x16/0x1b
kernel: [93272.427584]
kernel: [93272.427587] Leftover inexact backtrace:
kernel: [93272.427588]
oracleasm module를 unloading하는 과정에서 Superblock list가 메모리에 존재하지 않아, Superblock Callback 함수들을 Access 시도하는 과정에서 System crash 유발
-------------------------------------------------
SLES11SP2 panics on reboot after shutting down Oracle services
http://www.novell.com/support/kb/doc.php?id=7010465
Environment
SUSE Linux Enterprise Server 11 Service Pack 2
oracleasm-2.0.5-7.24.59
oracleasm-kmp-default-2.0.5_3.0.13_0.27-7.24.59
Situation
Upon shutdown or reboot, the system panics shortly after shutting down the Oracle database and the system shutdown hangs. The following kernel trace is written to the system log:
slab error in kmem_cache_destroy(): cache `asmdisk_cache': Can't free all objects
Pid: 7997, comm: modprobe Tainted: G X 3.0.26-0.7-default #1
Call Trace:
[<ffffffff810048b5>] dump_trace+0x75/0x300
[<ffffffff814401af>] dump_stack+0x69/0x6f
[<ffffffff8113e0b8>] kmem_cache_destroy+0xd8/0x120
[<ffffffffa02319ca>] exit_asmfs_fs+0x1a/0x650 [oracleasm]
[<ffffffff8109775e>] sys_delete_module+0x17e/0x270
[<ffffffff8144af92>] system_call_fastpath+0x16/0x1b
DWARF2 unwinder stuck at system_call_fastpath+0x16/0x1b
Cause
While unloading the oracleasm module asmfs is still in the list of superblocks but no longer available in memory causing the system to crash while trying to access the callback functions in the superblock.
Resolution
Please update to oracleasm-2.0.5-7.26.3 and oracleasm-kmp-default 2.0.5-7.26.3 or later.
-------------------------------------------------
Action Plan 1.
What: Oracleasm과 oracleasm-kmp-default를 Upgrade
Why : 예기치 않은 패닉 발생을 방지하기 위해
. oracleasm-2.0.5-7.26.3 or later
. oracleasm-kmp-default 2.0.5-7.26.3 or later