cancel
Showing results for 
Search instead for 
Did you mean: 

SCSI ERRORS

Yiwen
Level 3
Hi all

i did again all the installation and now i think is every things is ok but when i launch the backup from the SAN media servers (solaris 9) i receive this errors in the system logs

and the backup hang :

the bellow output from Activity monitor

09/02/2009 16:02:32 - requesting resource Minsatcs02-hcart-robot-tld-0
09/02/2009 16:02:32 - requesting resource netbackup.NBU_CLIENT.MAXJOBS.Minsatcs02
09/02/2009 16:02:32 - requesting resource netbackup.NBU_POLICY.MAXJOBS.minsat02-OS
09/02/2009 16:02:32 - Waiting for scan drive stop HP.ULTRIUM4-SCSI.000, Media server: Minsatcs02
09/02/2009 16:02:33 - granted resource  netbackup.NBU_CLIENT.MAXJOBS.Minsatcs02
09/02/2009 16:02:33 - granted resource  netbackup.NBU_POLICY.MAXJOBS.minsat02-OS
09/02/2009 16:02:33 - granted resource  A00003
09/02/2009 16:02:33 - granted resource  HP.ULTRIUM4-SCSI.000
09/02/2009 16:02:33 - granted resource  Minsatcs02-hcart-robot-tld-0
09/02/2009 16:02:33 - estimated 0 kbytes needed
09/02/2009 16:02:34 - started process bpbrm (pid=15440)
09/02/2009 16:02:34 - connecting
09/02/2009 16:02:34 - connected; connect time: 0:00:00
09/02/2009 16:02:37 - mounting A00003



and this from system messages:

[ID 832037 daemon.error] scsi command failed, may be timeout, scsi_pkt.us_reason = 6

is this becoz somes missed patches or others issues

i checked  the docs of this HBA qlogic and no specific patch are required for installation.
7 REPLIES 7

Rakesh_Khandelw
Level 6

Run sgscan tape or  scan -tape command and check if you can see tape drives or not.

I see you have LTO-4 drive but what make/model of tape library you have.

Omar_Villa
Level 6
Employee
SCSI Error can fall in many things, from physical issues to configuration, thinking on this last one, you need to check 3 things from the following commands outputs

vmoprcmd -h <media server> -shmdrive | awk '{print $38,$39,$31}'
vmoprcmd -h <media server> -autoconfig -t
vmglob -listall -java | grep -i <media server>


once you have this output you need to Match:

Serial Number
Drive Name
Path

The first command output will give you this 3 fields and is already sorted by Serial Number
Second command will only show Serial Number and Path
Third command shows Path and Drive Name

If they dont match in any way you will need to update the config with:

vmoprcmd -h <media server> -devconfig "-update -drive <idx> -path <path>"

after this update confirm you have a match running the commands again and do this with every drive that is not matching on any of the mentioned topics.


Hope this helps.
Regards.

Yiwen
Level 3
Thnx all

the problem that all the test they are ok , but is just the moment of the backup :

09/02/2009 16:02:34 - started process bpbrm (pid=15440)
09/02/2009 16:02:34 - connecting
09/02/2009 16:02:34 - connected; connect time: 0:00:00
09/02/2009 16:02:37 - mounting A00003


then netbackup hang in this step mounting mediaxxxx   and this message appear

[ID 832037 daemon.error] scsi command failed, may be timeout, scsi_pkt.us_reason = 6


and the links fail and sometimes the drive go down

and i reconfigure all the links again.


is normal to use LTO3 tapes for this TL wich is LTO4 drive???

bellow the commads requested:


1)
root@netbackup # sgscan tape
/dev/sg/c0tw500110a000946d5al0: (/dev/rmt/0): "HP      Ultrium 4-SCSI"

root@netbackup # vmoprcmd -h Minsatcs01 -shmdrive | awk '{print $38,$39,$31}'
HU183634L8 HP.ULTRIUM4-SCSI.000 /dev/rmt/0cbn
 
root@netbackup # vmoprcmd -h Minsatcs02 -shmdrive | awk '{print $38,$39,$31}'
HU183634L8 HP.ULTRIUM4-SCSI.000 /dev/rmt/1cbn

root@netbackup # vmoprcmd -h Minsatcs03 -shmdrive | awk '{print $38,$39,$31}'
HU183634L8 HP.ULTRIUM4-SCSI.000 /dev/rmt/1cbn

2)
root@netbackup # vmglob -listall -java | grep -i Minsatcs01
VMGLOB4.5 robot ROBOT0 0838BR0022 Minsatcs01 netbackup 0 -1 TLD - 0x80000 - 0 HP~~~~~~MSL~G3~Series~~~D.00 - - netbackup -1 -1 -1 -1
VMGLOB4.5 drive HP.ULTRIUM4-SCSI.000 HU183634L8 Minsatcs01 netbackup 0 1 TLD hcart 0x2000 - 0 HP~~~~~~Ultrium~4-SCSI~~H44W - - - -1 -1 -1 -1

root@netbackup #
vmglob -listall -java | grep -i Minsatcs02
VMGLOB4.5 robot ROBOT0 0838BR0022 Minsatcs02 netbackup 0 -1 TLD - 0x80000 - 0 HP~~~~~~MSL~G3~Series~~~D.00 - - netbackup -1 -1 -1 -1
VMGLOB4.5 drive HP.ULTRIUM4-SCSI.000 HU183634L8 Minsatcs02 netbackup 0 1 TLD hcart 0x2000 - 0 HP~~~~~~Ultrium~4-SCSI~~H44W - - - -1 -1 -1 -1

root@netbackup # vmglob -listall -java | grep -i Minsatcs03
VMGLOB4.5 robot ROBOT0 0838BR0022 Minsatcs03 netbackup 0 -1 TLD - 0x80000 - 0 HP~~~~~~MSL~G3~Series~~~D.00 - - netbackup -1 -1 -1 -1
VMGLOB4.5 drive HP.ULTRIUM4-SCSI.000 HU183634L8 Minsatcs03 netbackup 0 1 TLD hcart 0x2000 - 0 HP~~~~~~Ultrium~4-SCSI~~H44W - - - -1 -1 -1 -1

3///////////////
root@netbackup # vmoprcmd -h Minsatcs01 -autoconfig -t
TPAC60 HP      Ultrium 4-SCSI  H44W HU183634L8 -1 -1 -1 -1 /dev/rmt/0cbn - -
root@netbackup #
root@netbackup # vmoprcmd -h Minsatcs02 -autoconfig -t
TPAC60 HP      Ultrium 4-SCSI  H44W HU183634L8 -1 -1 -1 -1 /dev/rmt/1cbn - -
TPAC60 HP      C7438A          V309 0005476925 -1 -1 -1 -1 /dev/rmt/0cbn - -

root@netbackup # vmoprcmd -h Minsatcs03 -autoconfig -t
TPAC60 HP      Ultrium 4-SCSI  H44W HU183634L8 -1 -1 -1 -1 /dev/rmt/1cbn - -
TPAC60 HP      C7438A          ZP5A HU10629YD5 -1 -1 -1 -1 /dev/rmt/0cbn - -

even when i aply this command still the same.

vmoprcmd -h Minsatcs01 -devconfig "-update -drive 0 -path /dev/rmt/0cbn"


thnx a lot.


Yiwen
Level 3
hi every body

in addition of my questions above,
please someone can tel me when i have the bellow conflict how i can fix it and renew the EMM database,

even if i do the a new invotory  the errors still.

root@netbackup #
root@netbackup #
vmphyinv -rn 0
Waiting for mount of media ID A00000 on host netbackup
Mounted media ID A00000 on host netbackup
Waiting for mount of media ID A00001 on host netbackup
Mounted media ID A00001 on host netbackup
Waiting for mount of media ID A00002 on host netbackup
Mounted media ID A00002 on host netbackup
Waiting for mount of media ID A00003 on host netbackup
Mounted media ID A00003 on host netbackup
Waiting for mount of media ID A00004 on host netbackup
Mounted media ID A00004 on host netbackup
Waiting for mount of media ID A00005 on host netbackup
Mounted media ID A00005 on host netbackup
Waiting for mount of media ID A00006 on host netbackup
Mounted media ID A00006 on host netbackup

Proposed Change(s) to Update the Volume Configuration
=====================================================
Logically move media ID A00000 from slot 1 to standalone residence.
Logically move media ID A00002 to slot 1.
Logically move media ID A00001 from slot 2 to standalone residence.
Logically move media ID A00004 to slot 2.
Logically move media ID A00003 from slot 6 to standalone residence.
Logically move media ID A00001 to slot 6.
Logically move media ID A00000 to slot 9.

Following error(s) detected:

Media ID conflict detected: Media in slot 5 and 1 have the same media ID
A00002 on the tape header.
Media ID conflict detected: Media in slot 11 and 6 have the same media ID
A00001 on the tape header.

root@netbackup # robtest

thnx.

J_H_Is_gone
Level 6
Media ID conflict detected: Media in slot 5 and 1 have the same media ID
A00002 on the tape header.
Media ID conflict detected: Media in slot 11 and 6 have the same media ID
A00001 on the tape header.


do you have barcodes on the tapes?

When you have a barcode on the tape , the first time it is written to the barcode number is put on the header of the tape.    This is a safety issue so the if the barcodes get switched it and the barcode and header do not match it will not write to it.

If you do not have barcodes on the tapes then NB gives them names like A00001.

So you put the first tape in the library and it gets called A00001 and you write to it so it gets a header of A00001.  Now you remove that tape from the library and you delete it.

A new tape goes in the library NB no longer knows it has a tape called A00001 so this new tape gets called A00001 and when written to the header gets A00001 to it. 
Now you put the original A00001 into the library and NB checks it out and finds that it also as a header of A00001  - you now have a conflict.

you will have to bplabel or erase one of the tapes or remove one from the library to get rid of the conflict.

My first thought is that you do not have barcodes on your tapes, and NB can be very hard to use with out barcodes.

Omar_Villa
Level 6
Employee
the 3 DB's look good, check under the library for any errors in the logs, other thing that some times helps is to reset the HBA and look for any HBA issues under the box a good way to isolate this is to confirm if the error is over every media server or only on Minsatcs02, because if the drive is working for the other boxes than is a media server problem, if not, than is a drive issue, you can confirm any communication to drive with the command:

vmoprcmd -h <media server> -devconfig "-dev_ping -drive -path <path>"

Run this on every media server and see if the drive is available for all the medias.


Regards.

Anonymous
Not applicable
J,
Here was his previous post. Where he described at one point he removed the barcode labels!
https://www-secure.symantec.com/connect/forums/san-media-server

Using NetBackup without barcode labels is asking for trouble when it comes to protecting your data and knowing where it is for restores.

Advise: Erase all the tapes, and put barcode labels on them. Put in library, inventory it. Then bplabel the tapes or use the GUI right click and Label.

Tip: Assign a rule or configure library to read from left to right for first 6 chars so you have unique barcodes that do not include the generation of tape.
eg EZX943L4 barcode in Netbackup is EZX934