Reserve errors on shared drive b/w media servers and HNAS Filer

[Abhishek_Shar1]

Hi,

We are seeing a lot of SCSI reserve errors on a tape drive shared b/w media servers and a NDMP HNAS filer .The drive goes down and backups are failing

to these drive .We have recently shared this tape drive .Are there any prequesites for configuring this or I am missing something  .Scsi reservations

is set in Netbackup but i am not sure if its set on HNAS filer

Please see below errors in logs

Also the tape is mounted and then the drive fails to scsi reserve the drive it seems .

Tape drive detail

OY---hcart-005-ndmp      No      No                     No        hcart
    PNIOYNBMDS01               /dev/mt_d2l0 (pnievs1)               SCAN-TLD
    PNIOYNBMDS01               /dev/mt_d6l0 (pnipbevs20)            TLD
    PNIOYNBMDS01               /dev/mt_d2l0 (pnievs5)               TLD
    PNIOYNBMDS01               /dev/nst10                           TLD
    PNIOYNBMDS02               /dev/mt_d2l0 (pnievs1)               TLD
    PNIOYNBMDS02               /dev/mt_d6l0 (pnipbevs20)            TLD
    PNIOYNBMDS02               /dev/mt_d2l0 (pnievs5)               TLD
    PNIOYNBMDS02               /dev/nst8                            DOWN-TLD
    pnioynbvstg01              {4,0,4,0}                            DOWN-TLD
    pnioydmzmds1               /dev/nst4                            DOWN-TLD
    pnioynbvstg02              {4,0,0,0}                            ACTIVE
    pnioydmzmds2               /dev/nst4                            DOWN-TLD

BPTM log

03:44:31.141 [11834] <2> parse_resource_strings: VOLUME 1 RS1453 4017979 RS1453 NetBackup HP I120608442 6 8 0 299 0 {00000000-0000-0000-0000-

000000000000} 0
03:44:31.141 [11834] <2> parse_resource_strings: DRIVE 3 OY---hcart-005-ndmp 2000534 HUE1216F19 /dev/nst10 -1 -1 -1 -1 -1 -1 -1 -1 *NULL* *NULL* *NULL* *NULL* 1 0 0 1 0 0
03:44:31.156 [11834] <2> mount_open_media: Waiting for mount of media id RS1453 (copy 1) on server PNIOYNBMDS01.
03:44:31.156 [11834] <4> create_tpreq_file: symlink to path /dev/nst10
03:44:31.157 [11834] <2> manage_drive_before_load: retrying SCSI RESERVE (reserve unit scsi command failed, status = 0x18, key = 0x0, asc = 0x0, ascq = 0x0)
03:44:36.158 [11834] <2> manage_drive_before_load: retrying SCSI RESERVE (reserve unit scsi command failed, status = 0x18, key = 0x0, asc = 0x0, ascq = 0x0)
03:44:41.159 [11834] <2> manage_drive_before_load: retrying SCSI RESERVE (reserve unit scsi command failed, status = 0x18, key = 0x0, asc = 0x0, ascq = 0x0)
03:44:46.159 [11834] <2> manage_drive_before_load: retrying SCSI RESERVE (reserve unit scsi command failed, status = 0x18, key = 0x0, asc = 0x0, ascq = 0x0)
03:45:11.184 [11834] <2> log_media_error: successfully wrote to error file - 09/27/12 03:45:11 RS1453 -1 RESERVE_ERROR OY---hcart-005-ndmp 0 1 0 0
04:15:03.597 [13479] <16> really_tpunmount: error unloading media, TpErrno = Robot operation failed

OS logs

 Reservation Conflict status from OY---hcart-005-ndmp (device 5)
Sep 27 21:46:40 PNIOYNBMDS01 avrd[8678]: ndmp_tape_open failed on device 5, result is 2 (NDMP_DEVICE_BUSY_ERR)
Sep 27 21:46:56 PNIOYNBMDS01 avrd[8678]: ndmp_tape_open failed on device 5, result is 2 (NDMP_DEVICE_BUSY_ERR)
Sep 27 21:47:11 PNIOYNBMDS01 avrd[8678]: ndmp_tape_open failed on device 5, result is 2 (NDMP_DEVICE_BUSY_ERR)
Sep 27 21:47:26 PNIOYNBMDS01 avrd[8678]: ndmp_tape_open failed on device 5, result is 2 (NDMP_DEVICE_BUSY_ERR)

Media error logs

09/27/12 03:45:11 RS1453 -1 RESERVE_ERROR OY---hcart-005-ndmp 0 1 0 0
09/27/12 14:49:46 RS1900 -1 RESERVE_ERROR OY---hcart-005-ndmp 0 1 0 0
09/27/12 18:05:50 RS1374 -1 RESERVE_ERROR OY---hcart-005-ndmp 0 1 0 0
09/28/12 01:50:48 RS1374 -1 RESERVE_ERROR OY---hcart-005-ndmp 0 1 0 0

Thanks,

Abhishek

[Douglas_A]

It sounds like your NAS does not support the SCSI reservation method that NBU is using..

The solution is to either configure SCSI-2 reserve/release on the Filer or to configure persistent reservation within Netbackup.

If you want to configure Persistent reservation in NBU, open the Host Properties of the Master > Properterties on master server > Media and modify the SCSI reservation option. Then restart NBU services and see if the error persists.

[Yasuhisa_Ishika]

Hitachi NAS(BlueArc Titan) does not support SCSI reservation by default. You need to enable it.
Please check this thread.

https://www-secure.symantec.com/connect/forums/ndmp-job-failing-when-reaches-tape-fragmentation-size

BTW, if you can not manage it anyway, consider to disable SCSI reservation feature in NetBackup.
I believe there are no need to use SCSI reservation, in general, because resources including tape drives are managed by nbrb, and it can not occur that multiple host under NetBackup control access the drive simultaneously.

[Abhishek_Shar1]

Hi,

Thanks guys for your help .We are working with the Storage team to check the option if they can enable 

SCSi reservation on the Filer side as it is disabled right now else we will disable it from Netbackup end and see if the issue gets resolved . 

Abhishek