Duplication failed because physical go down continuosly

[Wolaf25]

Hi team

I have problem with the duplication with SLP, because the drives continuosly go down, and if the drives are up, the SLP dont use all of them. Also, if I run manual duplicattion`s, they run good....

The master, the Data Domain, the physical libraries are under FC SAN, are joined with a FC switch.
 

Netbackup is 7.5.0.4, the master is installed in linux Redhat Red Hat Enterprise Linux Server release 6.7 (Santiago).

The master is HP ProLiant - DL380p Gen8 –

RAM :                                                    

free

                     total            used                 free        shared      buffers     cached

Mem:     132119292  128450800    3668492     134052     729868  123522632

Drivers information

Netbackup side

/opt/openv/volmgr/bin/tpautoconf -t

TPAC60 IBM     ULT3580-HH5     E6Q3 1068206474 -1 -1 -1 -1 /dev/nst41 - -

TPAC60 IBM     ULT3580-HH5     E6Q3 1068206586 -1 -1 -1 -1 /dev/nst40 - -

The backups are using Data Domain Model: DD860, OS: 5.4.4.2-453843, VTL mode.

Aditional we have others medias, with UNIX 7.1 and 6.1 SO and linux Redhat too.

Also the master have  1 physical tape library, TS3100, with 2 physical drives.

Product ID

3573-TL

Currently Installed Library Firmware	C.50 / 3.20e
Bootcode Firmware Revision	0.80
Barcode Reader	SE625
IPv4 Address	163.35.7.72
MAC Address	000E111597D8
Library Mode	Random
WWide Node Name	2000000E111597D8

Drives :

Vendor ID	IBM
Product ID	ULT3580-HH5
Serial Number	1068202992
Firmware Revision	E6Q3
Element Address	257
Control Path Drive	No
Data Compression	Yes
Interface Type	Fibre Channel
Node Name	2004000E111597D8
Port A	Enabled
Port Name	2005000E111597D8
Topology	N-Port
Speed	Automatic
Port B	Disabled

I run robtest with option s d, and all look well.

When in Netbackup GUI I saw the drivers down, I run the following process, in the servers,  for Up the drives

1.- /usr/bin/rescan-scsi-bus.sh

2.- /opt/openv/volmgr/bin/tpautoconf -a

or   /opt/openv/volmgr/bin/tpconfig -d

 sometime the drives go UP, in others no, I need to restart Netbackup service or in worst case I need reboot the server.

What happen here ??

If you need any other information, please, ask to me.

Thanks for your support.

[Will_Restore]

Library firmware very old (2014), way past time for an update.  There is also recent update for the tape drives.  Call the vendor. 

[Michael_G_Ander]

Know there has been a problem on some HP LTO3 tape drives, where the tape drives lost the login to the fabric wonder if the IBM tape drives could suffer from a similar problem.

Another thing that could cause this is changes in the SAN, if persistent binding is not used on the tape drives and library.

Regarding the SLP, check the number of concurrent write drives on the recieving storage unit.

[pats_729]

Please check /usr/openv/netbackup/db/media/errors file for further error related Drive and Tapes. If you see hardware related errors, a call to hardware vendor is a must.

[Wolaf25]

Hi

Will restore

The library firmware, I'm not sure if I can upgrade, because, we have some internal regulations, one sector, who decide if we can upgrading or not the firmware. I must investigate on internal way.

Michael, aparently the problem is only with duplications with SLP, if I use manual duplication they run well. When are running duplications SLP, not use all the drives. I see drives idle but, the duplications are in  queue with the messages, there are not drives free .... DO you know if  is an internal limitations for the SLP duplications runing at the same time? Some parameters or configurations?

SP.NBU

I check this file and I see this messages

04/07/16 12:11:04 A51431 41 WRITE_ERROR IBM.ULT3580-HH5.001

What its mean?

Thanks for your cooperations.

[pats_729]

Is that the same time when duplication failed / drive went down ? If so then Probably drives required cleaning or really need a check up by vendor.

[Wolaf25 · Accepted solution]

Hi Guys

The issue was resolved, aparently was caused for a low space in /opt/openv, the fs was full and I added more space, reestarted backup service, and these extranges problems dessapear...

Now, manual duplications and SLP's duplications run well.

But I have another issue now.. Will open other discussion.

Thanks for your cooperations!!!