00000300abc34480 300aaf7f318 60044ef57c8 1 59 300afce9430 PC: cv_wait_sig+0x114 CMD: mount /dev/vx/dsk/sharedg/transferlv /var/transfer stack pointer for thread 300abc34480: 2a106ac4961 [ 000002a106ac4961 cv_wait_sig+0x114() ] vx_event_waitsig+0x28() kdm_eventprocess+0x54() kdm_eventpost+0x1f0() kdm_mount_event_gen+0x218() vx_mount_dmsetup+0x1b4() vx_mount+0x760() domount+0x9d0() mount+0x108() syscall_ap+0x58() syscall_trap32+0xcc()5612/1: 1.1620 0.0001 open64("/dev/vx/rdsk/sharedg/transferlv", O_RDONLY) = 7
5612/1: 1.1621 0.0001 ioctl(7, (('V'<<24)|('O'<<16)|('L'<<8)|2), 0xFFBFF4E0) = 0
5612/1: 1.1979 0.0358 close(5) = 0
5612/1: 1.1981 0.0002 ioctl(7, (('V'<<24)|('O'<<16)|('L'<<8)|30), 0x001E03B0) Err#25 ENOTTY
5612/1: 1.1985 0.0004 pread64(7, "A501FCF5\0\0\007\0\004\0".., 8192, 8192) = 8192
5612/1: 1.1990 0.0005 fdsync(7, FSYNC) = 0
5612/1: 1.1991 0.0001 close(7) = 0
5612/1: 1.1992 0.0001 stat64("/var/tmp/test1", 0xFFBFFB00) = 0 <== Will hang after hereWith IBM Tivoli Storage Manager HSM Client package (TIVsmChsm), a possible deadlock situation in DMAPI can cause mount commands to hang. VxFS File systems mounts without any problem through vfstab because they get mounted before the TSM HSM client starts up. The /etc/rc2.d/S98adsmhsm script starts up TSM HSM before VCS startup script start so this affects mounting of VxFS file systems through VCS.
The issue was noticed with Tivoli Storage Manager package (TIVsmCapi) version 5 Release 5 Level 2.0 and upgrading this package version to 6 seems to solve this issue. Please check with IBM for supportability and upgrade of the Tivoli Storage Manager.
As a workaround, HSM client can be stopped or disabled by renaming /etc/rc2.d/S98adsmhsm script.