device locked

[rischard75]

Hi,
my configuration: NT5.0-DataProtector-IBM LTO 3583 2drivers

Sometimes stays my device locked after backup session and when starts the next backup, it cannot write to device (Busy Drive Handling = Eject). The process bma.exe (media agent client) still run after uncessfully backup session. Data Protector don't stop it I must do it manually. Can be the reason of different versions of installed Data Protector clients (all are windows clients)? Or some of them are patched and the other not?

I've found this in the debug.txt:

15/04/2003 9:00:52 BSM.1012.1060 ["lib/ipc/ipc.c /main/r50_fix/2":423] A.05.00 bDPWIN_00005
IpcCloseConnection: IpcHandle -1 out of range (0..200)
Function trace (currentLevel=1):
Level 1: IpcGetEntry()
Level 0: IpcCloseConnection()

4/15/2003 9:01:11 AM BMA.3276.2792 ["ma/spt/sctl_NT.c /main/101":631] A.05.00 b190
CreateOB2Device: SctlDevLock error [Mutex already locked.
System error: [122] The data area passed to a system call is too small.]

15/04/2003 9:14:36 MMD.3320.3040 ["sm/mmd/mmdquery.c /main/94":744] A.05.00 b190
[MsgLockDeviceList] could not unlock device LTOdrv3 by req. from host xxx
SctlDevLock: Request for mutex "OmniBack1:5:0:0:2" failed (scsi5:0:0:2 - type 1 err=122)

What do the messages mean?

[Saulius_1]

Actually I don't know what debug log means, but, I can tell you some tips from my personal expierence:

1. It's better to fight the reason rather than consequences. From my expierence normally configured omniback backup stops in three cases

a) Too many files in the folder (directory). Solution - exclude this directory from backup, or reduce number of files in it. I think there is a possibility to increase timeouts somewhere, but I don't know where.

b) Network problems, like packet loss, insufficient bandwidth (router can not handle it, and drops packets, etc.). If network connection breaks, it seems that Omniback can fell into deep meditation phase. These problems should be solved in any case.

c) Old (too new) omniback agent. I've seen strange behaviour with integrated Omniback agents on HP NAS8000 for example. I don't know any other sollution to this problem than stell patience ...

2. Every morning check backup logs, and if something wrong run:

omnidbutil -show_locked_devs

if something stayed locked issue:

omnidbutil -free_locked_devs [ devname | mediumId | cartName phyLocation ]

Hope it will help.

Saulius