problem starting cloning

[Robert800]

Have problem starting a cloning operation it just mount a recycable tape and after that it read on the tape but then it just unmount the recycable tape and tries another one without any luck. I can do restore from tapes but cloning doesnt seems to work, i have also done a cleaning tape.

The only thing i see in the log is this: 3 0 5 media 0 7 warning 0 56 \\.\Tape2 moving: fsf 179: No more data is on the tape.

C:\Documents and Settings\adm>inquire
scsidev@0.0.0:HL-DT-STCD-ROM GCR-8240N1.10|CD-ROM
scsidev@2.0.0:HP Ultrium 4-SCSI H46Z|Tape, \\.\Tape0
S/N: F09EE64000
ATNN=HP Ultrium 4-SCSI F09EE640
00
WWNN=500308C09EE64000
WWPN=500308C09EE64001
WW2N=500308C09EE64000
PORT=00000001
scsidev@2.0.1:ADIC Scalar i500 520G|Autochanger (Jukebox), \\.\Changer0
S/N: ADICA0C0108A20_LLA
ATNN=ADIC A0C0108A20_LLA

scsidev@2.1.0:HP Ultrium 4-SCSI H46Z|Tape, \\.\Tape1
S/N: F09EE64004
ATNN=HP Ultrium 4-SCSI F09EE640
04
WWNN=500308C09EE64004
WWPN=500308C09EE64005
WW2N=500308C09EE64004
PORT=00000001
scsidev@2.2.0:HP Ultrium 4-SCSI H46Z|Tape, \\.\Tape2
S/N: F09EE64090
ATNN=HP Ultrium 4-SCSI F09EE640
90
WWNN=500308C09EE64090
WWPN=500308C09EE64091
WW2N=500308C09EE64090
PORT=00000001
scsidev@10.0.1:IFT A24F-G2430 361C|Disk
S/N: 01B1F3000000004838B75C000001
WWNN=200000D02301B1F3
scsidev@10.2.0:IFT A24F-G2224-1 347B|Disk
S/N: 0DD10D4649A5B300
WWNN=200000D0230DD10D
scsidev@10.2.1:IFT A24F-G2224-1 347B|Disk
S/N: 0DD10D61D2AD1100
WWNN=200000D0230DD10D
scsidev@12.6.0E/PV 1x8 SCSI BP 1.0 |Processor
scsidev@14.57.0: RAID DummyDevice 0000|Unknown Device Type
scsidev@15.0.0ERC LD 0 PERCRAID |Disk



deamon.raw:
33526 1237828227 2 0 0 5072 4404 0 backup nsrmmd 33 Start %s, with PID %d, at HOST %s 3 20 10 nsrmmd #10 1 4 4404 12 25 backup
38752 1237828229 2 0 0 2676 2672 0 backup nsrd 5 %s %s 2 21 9 \\.\Tape2 49 45 38337 24 %s operation in progress 1 0 5 Eject
67986 1237828304 2 0 0 2532 2168 0 backup nsrmmgd 50 Unloading volume `%s' from device `%s' to slot %d. 3 22 8 000452L3 21 9 \\.\Tape2 1 3 115
67985 1237828324 2 0 0 2532 2168 0 backup nsrmmgd 52 Loading volume `%s' from slot `%d' into device `%s'. 3 22 8 000447L3 1 2 61 21 9 \\.\Tape2
38752 1237828346 2 0 0 2676 2672 0 backup nsrd 5 %s %s 2 21 9 \\.\Tape2 49 53 38337 24 %s operation in progress 1 0 12 Verify label
38752 1237828360 2 0 0 2676 2672 0 backup nsrd 5 %s %s 2 21 9 \\.\Tape2 49 45 38337 24 %s operation in progress 1 0 5 Mount
38758 1237828419 2 0 0 2676 2672 0 backup nsrd 9 %s %s: %s 3 0 5 media 0 7 warning 0 55 \\.\Tape2 moving: fsf 39: No more data is on the tape.
38758 1237828509 2 0 0 2676 2672 0 backup nsrd 9 %s %s: %s 3 0 5 media 0 7 warning 0 55 \\.\Tape2 moving: fsf 39: No more data is on the tape.
38758 1237828600 2 0 0 2676 2672 0 backup nsrd 9 %s %s: %s 3 0 5 media 0 7 warning 0 55 \\.\Tape2 moving: fsf 39: No more data is on the tape.
42506 1237828600 2 0 0 2676 2672 0 backup nsrd 9 %s %s: %s 3 0 5 media 0 4 info 49 64 9055 11 fsf %ld: %s 2 2 2 39 0 30 No more data is on the tape.

42506 1237828600 2 0 0 2676 2672 0 backup nsrd 9 %s %s: %s 3 0 16 write completion 0 6 notice 49 55 67896 30 Writing to volume %s completed 1 22 8 000447L3
12361 1237828604 2 0 0 2676 2672 0 backup nsrd 27 %s Finished with status: %s 2 0 35 [Jukebox `Bender', operation # 30]. 0 9 succeeded
0 1237828610 2 0 0 2676 2672 0 backup nsrd 2 %s 1 0 70 Operation 31 started : Load volume `A00022L4', volume id `197022538'..
38752 1237828612 2 0 0 2676 2672 0 backup nsrd 5 %s %s 2 21 9 \\.\Tape2 49 45 38337 24 %s operation in progress 1 0 5 Eject
67986 1237828680 2 0 0 2532 2168 0 backup nsrmmgd 50 Unloading volume `%s' from device `%s' to slot %d. 3 22 8 000447L3 21 9 \\.\Tape2 1 2 61
67985 1237828702 2 0 0 2532 2168 0 backup nsrmmgd 52 Loading volume `%s' from slot `%d' into device `%s'. 3 22 8 A00022L4 1 1 7 21 9 \\.\Tape2
38752 1237828726 2 0 0 2676 2672 0 backup nsrd 5 %s %s 2 21 9 \\.\Tape2 49 53 38337 24 %s operation in progress 1 0 12 Verify label
38752 1237828741 2 0 0 2676 2672 0 backup nsrd 5 %s %s 2 21 9 \\.\Tape2 49 45 38337 24 %s operation in progress 1 0 5 Mount
38758 1237828863 2 0 0 2676 2672 0 backup nsrd 9 %s %s: %s 3 0 5 media 0 7 warning 0 56 \\.\Tape2 moving: fsf 179: No more data is on the tape.

Has anyone any tip becouse iam out of ideas on this one, we are using NW 7.5.

 

[605]

It's probably worth to check if the statement is true. Use "scanner -m vv device"to check whether there is really no more data on tape.

 

[a060463xyz]

With NW 7.4.3 and newer I could not clone saveset with status recyclable.
I got entries in the daemon.raw like the following :

39078 26.03.2009 14:48:15 0 0 2 4180 1560 0 <myservername> nsrmmdbd Media DB error: Save set ssid:742421008 cloneid:1238074208 eligibility can not be cleared, retention time must be adjusted first

After changing the retentiontime of the saveset with nsrmm -e "1 day" -S <ssid>/<cloneid>
everything worked for me.
So I think its not a bug but a new feature

 

[Robert800]

Thanks but it not exactly the same problem, the problem is that when we clone it writes to a tape for example 5 GB then it says the tape is full and searches for the next "recyclable" tape and so on. then it for example writes 100 GB on the next tape then it say that the tape is full.

I did a scanner -m \\.\Tape2 on a tape like this and it said the same wrong size and that the tape is full.

On the other hand we have some tapes that write normal when we start clone.

We have both LTO3 and LTO4 tapes but the same problem occur on both tapes.

 

[a060463xyz]

Sorry i did not read carefully enough, i thought you want to clone recyclable tapes.
After rereading carefully ;-) i think you have Fibrechannel issues.
Can you separate the Tapes on different FC fabrics ?
e.g reading from Tape0 using the first fc interface of your NW Server and writing to Tape1 using the second fc interface ?
Two LTO-4 can saturate a 2 GB FC interface easily and a 4 GB FC link nearly.
Check your LTO-drivers. Dont use HP drivers, use IBM ATape or the Microsoft native LTO drivers instead.
Check the configuration of the Tapes and the Fc hbas. In most cases the factury defaults are "loop prefered", change it to "point-to-point".

 

[Robert800]

The problem started when we upgrade to Networker 7.5 and on a new server (64 bit win2k3)) and a new robot i500. We moved the old HBA card from the old server to the new server so we would have the same WWN names so we didnt had to do that again.

Everything seemed to work fine the first 2-4 weeks then it started to dont fill the tapes.

We use the HP drivers today becouse we havent found any IBM ATape or microsoft native LTO drivers, were can i find them?

 

[swekakadm]

Hi,
This is another way to describe the same problem that my colleague has describe it:

Legato don't fillout tapes

--------------------------------------------------------------------------------

Hi,

We have a EMC Legato NetWorker 7.5 Build 179 on Windows Server 2003 x64 DELL Server.
We do also have a Quantum i500 with 3x HP LTO-4 tapes connect with fiber to a QLogic QLA2462 PCI-X, the PCI-X bus is only a 133 Mhz bus. and the Disk we have is split up in 3 different LUNs connected via Fiber to a QLE2460 on PCI Express x4 bus (2,5 Ghz).

When Legato works we are only get 30 MB/s to our LTO-4 drivers. And normally in I see in TSM at least 100 MB/s. But I think this is a fragmentation problem that Legato creates on our Diskpool.

The issue we got is when we try to move data from Disk to Tape, then is Legato only filling the tapes randomize. We can see tapes that have a couple of MB to max size of 800 GB / tape.

The only thing we see is comment when Legato can't fill the tapes is, we got an error in Daemon.raw that say

38758 1238138368 2 0 0 2488 2484 0 SERVERNAME nsrd 9 %s %s: %s 3 0 5 media 0 7 warning 0 86 \\.\Tape1 writing: The request could not be performed because of an I/O device error.
42506 1238138368 2 0 0 2488 2484 0 SERVERNAME nsrd 9 %s %s: %s 3 0 5 media 0 6 notice 49 89 9142 22 %s %s %s on %s is full 4 0 13 LTO Ultrium-4 0 4 tape 22 8 000395L3 21 9 \\.\Tape1

Now to my questions and what I have done.
* The tapes that Legato tells me it is full but have only move less 100GB to and the Save Set has status Aborted. Can I somehow make an AUDIT VOLUME on that tape? Or can I delete that tape and retry the operation?

* We have moved the tapes from one HBA Emulex to a QLogic and it didn't help.

* If I only have fill out the tape to 50% but Legato tells me the tape is FULL and 10% of the Save Sets is Browsable, Can I a Run a MOVE CLIENT DATA from one tape to another so I can clean that tape and reuse it again?

* I sow someone tell us to use IBM Drivers for our HP drivers? The PCI ID is not the same so Windows will not let us install IBM drivers for HP drivers. And I don't think Microsoft has any LTO-4 drivers for Windows.

Please if you have any other ideas how we can continue, I'm more then happy.

PS: If we do a Clone of a tape to tape do we have a read/write speed of 100-150 MB/s.
One idea I think it could be is that the DELL PCI-X bus can't handle the I/O that we try to send. So we got a timeout issue of the data?

 

[swekakadm]

Hi,
We are going to try to switch disk and tape so tape will use the fast HBA and Disk the slower HBA so the tapes will not be the slow part and hopefully get away from tape I/O retries. (if we got any...)

 

[a060463xyz]

use this Link for LTO-4

http://download.cnet.com/IBM-ULTRIUM-HH4-TAPE-DRIVE/3000-2122_4-175945.html

The Windows drivers are part of the installation :
select your Tape drive in the Hardware Manager.
Select Update Driver.
Dont conncet to the web, select "Install from a list...(Advanced)
select "Dont Search. I will choose ..."
uncheck "Show compatible Hardware"
Now you will see "LTO" choose this despite the warnings.

 

[issen007]

Hi,
You can download the latest IBM Tape driver from
ftp://ftp.software.ibm.com/storage/devdrvr/Windows/Win2003/Latest
And run
>install_nonexclusive.exe -t -f -r

If you are using TSM do you need to run
>install_exclusive.exe -t -f -r

You can't use the IBM driver because you have HP LTO-4 Drives and not IBM LTO-4 drives. If you want the IBM driver to work with HP LTO-4 tapes drives do you need to modify the .INF file and insert the correct PCI value...

Try to archive Deomon.raw file once a day and see if all drives reports an I/O Device error or if only one drive reports it.
I know the log is hard to understand and no timestamps is in their.

Cheers
Christian

 

[swekakadm]

Seems we found the problem and it was the \\.\Tape1 that didnt write enough to the tapes so we have put it in service mode and made a call to Quantum about change it.

cheers