Restore old LTO4 after upgrade to LTO6 library

[Alessandro_Mend]

Hello,

I have changed last week the robotic and tape unit from LTO4 to LTO6, LTO4 labels range stop to C20140L4 and LTO6 label start from C20200L6 

I have unloaded all tapes LTO4 from the robotic and then removed old Device/Drives & Robot and then configured the new robot with "Device Configuration Wizard", I have 

So this is the situation now is this:

New LTO6 tape are right in inventory and they are correctly picked up from scratch pool.

Now I want to try to restore some data from LTO4 tapes. I have loaded into robotic but when I try to inventory I get this error:

Of course C20132 and C20133 are already in inventory as HCART (used with the old robotic) but assigned to the volume group "---" which means to me is outside of robotic.

I have tryed to change the media type to '1/2" cartridge type' into Advanced Robot Inventory during the inventory of LTO4 in order to make the same type but error is the same.

What is the right procedure to manage old LTO4 tapes and restore data from LTO4 tape using the new LTO6 drivers and robotic?

[Will_Restore]

change the old media type via command line:

<installpath>\volmgr\bin\vmchange  -new_mt hcart3  -m C20132

<installpath>\volmgr\bin\vmchange  -new_mt hcart3  -m C20135

either that or change the new drive type, but that's more troublesome in my book

[Genericus]

You need to set up bar code rules so one hcart value is for LTO6 and one is for LTO4. 

You will need to set one of your LTO6 drives (or a LTO4 if you kept it) as LTO4 hcart type.

I do not believe you can use the same hcart value for both LTO4 and LTO6

[Genericus]

Also - did you bypass the robot and basically open it up and take all the tapes out? 

If so, you likely need to reinventory it, and mark the tapes you took out as in what ever location you are storing them.

Run an audit of the robot, and then an inventory in netbackup.

[Marianne]

Do you have L4 and L6 media with same leading C20xxx characters? If so, it won't work. As you are adding the 1st 6 characters as media-id, having C20132L4 as media-id C20132 and C20132L6 also as same media-id is not possible. Please show us output of vmquery -m C20132 *** Edit *** Apologies... Reading your opening post again I see that your labels are not overlapping. Please show us vmquery output as the output will hopefully point us in the right direction.

[Genericus]

Can he change the robot and bar code rules to use 8 characters?

Otherwise it is better to get new codes for the LTO6 tapes, do not use C as 1st char

[mph999]

The media is being identified with barcode C20132L4 giving media id C20132 - note it says ' new media'   - this then failes with duplicate media ID.

So, in NBU you will see a media with C20132, what is the barcode showed along side it ??? Is it by anychance different than C20132L4 

If the barcode changes, for example to C20132, as the linbrary config has changed to display only the first 6 characters to NBU, then this will be identified as a 'new' tape, because the bardcode is different (C20132L4 is differnt that C20132, despite being the same physical tape).  However, as the meida ID is the using the first x6 characters, the media ID stays the same.  Hence the error, trying to add new tape, but getting duplicate media ID

Barcode   /  Media ID

C20132L4   / C20132   <<< Original

C20132 / C20132  <<< New, as barcode different, but mediaID the same

I suspect this is what has happened.

vmquery -m C20132 output should tell us, if you can post that up.

[Alessandro_Mend]

Thank you Marianne,

here the output of vmquery -m C20132:

================================================================================
media ID:              C20132
media type:            1/2" cartridge tape (6)
barcode:               C20132
media description:     ---
volume pool:           VP-05WEEKS-EXT (9)
robot type:            NONE - Not Robotic (0)
volume group:          ---
vault name:            ---
vault sent date:       ---
vault return date:     ---
vault slot:            ---
vault session id:      ---
vault container id:    -
created:               26/10/2015 10:07:27
assigned:              27/10/2015 20:39:21
last mounted:          31/10/2015 17:06:28
first mount:           27/10/2015 20:40:25
expiration date:       ---
number of mounts:      35
max mounts allowed:    ---
status:                0x0
================================================================================

[Alessandro_Mend]

Hello,

library is set to expose to NBU 8 digit from label

Here the output of 'robtest' (empty slot removed)

Opening {5,0,0,1}

MODE_SENSE complete

Enter tld commands (? returns help information)

s s

slot 1 (addr 4096) contains Cartridge = yes

Source address = 4096

Barcode = C20212L6

slot 2 (addr 4097) contains Cartridge = yes

Source address = 4097

Barcode = C20205L6

slot 3 (addr 4098) contains Cartridge = yes

Source address = 4098

Barcode = C20222L6

slot 4 (addr 4099) contains Cartridge = yes

Source address = 4099

Barcode = C20210L6

slot 14 (addr 4109) contains Cartridge = yes

Source address = 4109

Barcode = C20220L6

slot 16 (addr 4111) contains Cartridge = yes

Source address = 4111

Barcode = C20211L6

slot 18 (addr 4113) contains Cartridge = yes

Source address = 4113

Barcode = C20218L6

slot 19 (addr 4114) contains Cartridge = yes

Source address = 4114

Barcode = C20217L6

slot 20 (addr 4115) contains Cartridge = yes

Source address = 4115

Barcode = C20215L6

slot 22 (addr 4117) contains Cartridge = yes

Source address = 4117

Barcode = C20204L6

slot 23 (addr 4118) contains Cartridge = yes

Source address = 4118

Barcode = C20201L6

slot 24 (addr 4119) contains Cartridge = yes

Source address = 4119

Barcode = C20132L4

slot 25 (addr 4120) contains Cartridge = yes

Source address = 4120

Barcode = C20133L4

slot 34 (addr 4129) contains Cartridge = yes

Source address = 4129

Barcode = C20206L6

slot 35 (addr 4130) contains Cartridge = yes

Source address = 4130

Barcode = C20207L6

slot 36 (addr 4131) contains Cartridge = yes

Source address = 4131

Barcode = C20202L6

slot 37 (addr 4132) contains Cartridge = yes

Source address = 4132

Barcode = C20200L6

slot 38 (addr 4133) contains Cartridge = yes

Source address = 4133

Barcode = C20214L6

slot 39 (addr 4134) contains Cartridge = yes

Source address = 4134

Barcode = C20213L6

slot 40 (addr 4135) contains Cartridge = yes

Source address = 4135

Barcode = C20209L6

slot 41 (addr 4136) contains Cartridge = yes

Source address = 4136

Barcode = C20208L6

slot 42 (addr 4137) contains Cartridge = yes

Source address = 4137

Barcode = C20225L6

slot 43 (addr 4138) contains Cartridge = yes

Source address = 4138

Barcode = C20224L6

slot 44 (addr 4139) contains Cartridge = yes

Source address = 4139

Barcode = C20221L6

READ_ELEMENT_STATUS complete

..and here the output of 'vmquery -m C20132'

================================================================================
media ID:              C20132
media type:            1/2" cartridge tape (6)
barcode:               C20132
media description:     ---
volume pool:           VP-05WEEKS-EXT (9)
robot type:            NONE - Not Robotic (0)
volume group:          ---
vault name:            ---
vault sent date:       ---
vault return date:     ---
vault slot:            ---
vault session id:      ---
vault container id:    -
created:               26/10/2015 10:07:27
assigned:              27/10/2015 20:39:21
last mounted:          31/10/2015 17:06:28
first mount:           27/10/2015 20:40:25
expiration date:       ---
number of mounts:      35
max mounts allowed:    ---
status:                0x0
================================================================================

I suppose that the problem is 'that is found same media ID but for a different media type'...

[Marianne]

barcode:               C20132 

The problem is that the new library reads all 8 characters :  C20132L4

You will have to use vmchange command to update the barcode for each of the old LTO4 media-id's:

vmchange -barcode C20132L4 -m C20132 

You should be able to script this in a 'for' loop.
 

[Alessandro_Mend]

Thank you Marianne!

I have updated the media-id's and now I am able to inventory those tapes.

But now there is a second problem... when try to restore the data NBU seem unable to locate a right drive for restore.

How can I teach NBU that a drive type HCART3 may be able to read (and, obviously, only read) the tape of HCART type?

[Alessandro_Mend]

(I lost my previous post)

Thank you Marianne,

now the inventory was successful, but the restore was not succeeded.

During restore NBU seem that looking for HCART drives, how can I tell to NBU that HCART3 drives can be used to read (and only read) HCART media type?

[Marianne]

You need to change at least one LTO6 drive to hcart density to perform restores of hcart tapes. You can change it back to hcart3 after the restore. Although LTO6 drive can read LTO4 media, NBU has no idea about this underlying hardware compatibility. In NBU densities MUST match.

[Genericus]

I think you have to change the hcart type of one of the drives, to match the tape.

As I said above:

You will need to set one of your LTO6 drives (or a LTO4 if you kept it) as LTO4 hcart type.

[Will_Restore]

See my post above - change the media type, easier than changing the drive type.

[mph999]

No - at this poiunt, the media type isn't important.

THisissue, as pointed out below, is as I explained above, the barcode has changed, so NBU sees this as a new tape and tries to add it, but due to the media ID rules, the media ID works out to be the same as one already there.

[Marianne · Accepted solution]

To summarise : You had 2 problems: 1) Inventory because barcode in the NBU Volume database had 6 characters but the new robot sees 8 and caused the 'not unique' error. This was fixed with vmchange command. 2) Restore is an issue because the density of LTO4 tapes and LTO6 are different. I prefer to change the density of a tape drive for the duration of the restore. There have been suggestions to change the density of the media. This will also work but NBU may now try to use the LTO4 tapes for backups, which will fail as LTO6 cannot write to LTO4. Rather keep LTO4 as different density to keep it separate and change tape drive density for restores. I have been able to change tape drive density without restarting Device Manager service.

[Alessandro_Mend]

Thank you Marianne and thank you to all who have helped me to understand the problem.