Hi,
Maybe you would like to consider reconfiguring the everything, I'm listing
the steps here. Hope that helps.
1. FC-to-SCSI router
--------------------
*Not too sure which model you're using, how did you connect the 6 drives
(and robotic control)? Daisy-chained???
Quoting example for STK SN3250 (which only have 4 SCSI ports), the FC to
SCSI mapping table should look something like this for drives with robotic
control number set to 0 and drive SCSI ID 1,2,3,4,5,6 (configured using the
L180 panel)
I'm assuming you daisy-chained as follows:
Bus0, Port 0: Robot
Bus0, Port 1: drive 1 and drive 2
Bus1, Port 0: drive 3 and drive 4
Bus1, Port 1: drive 5 and drive 6
FC_LUN BUS TGT LUN
0x00 -> 0 0 0
0x01 -> 0 1 0
0x02 -> 0 2 0
0x03 -> 1 3 0
0x04 -> 1 4 0
0x05 -> 1 5 0
0x06 -> 1 6 0
0x05 -> - - -
0x06 -> - - -
0x07 -> - - -
0x08 -> - - -
...
...
2. HBA config file
------------------
Perform persistent binding for the router's wwn.
3. /kernel/drv/st.conf
----------------------
Add 9840 parameters to define the tape drive.
name="st" class="scsi" target=x lun=1;
name="st" class="scsi" target=x lun=2;
name="st" class="scsi" target=x lun=3;
name="st" class="scsi" target=x lun=4;
name="st" class="scsi" target=x lun=5;
name="st" class="scsi" target=x lun=6;
(* The target number depends on which target you bind to in your HBA config
file.)
4. Reboot
---------
# rm /dev/rmt/* to cleanup the device special files.
# touch /reconfigure
# sync; sync
# init 6
5. After reboot
---------------
Check dmesg or /var/adm/messages to verify that all drives are detected
6. NBU
------
/usr/openv/volmgr/bin/driver/sg.conf configuration
name="sg" class="scsi" target=x lun=1;
name="sg" class="scsi" target=x lun=2;
name="sg" class="scsi" target=x lun=3;
name="sg" class="scsi" target=x lun=4;
name="sg" class="scsi" target=x lun=5;
name="sg" class="scsi" target=x lun=6;
/usr/openv/volmgr/bin/driver/sg.links configuration
type=ddi_pseudo;name=sg;addr=x,1; sg/c\N0txl1
type=ddi_pseudo;name=sg;addr=x,2; sg/c\N0txl2
type=ddi_pseudo;name=sg;addr=x,3; sg/c\N0txl3
type=ddi_pseudo;name=sg;addr=x,4; sg/c\N0txl4
type=ddi_pseudo;name=sg;addr=x,5; sg/c\N0txl5
type=ddi_pseudo;name=sg;addr=x,6; sg/c\N0txl6
# mv /kernel/drv/sg.conf /kernel/drv/sg.conf.old
# /usr/openv/volmgr/bin/driver/sg.install
# sgscan all.
Over at NetBackup, configure the drive as hcart (1/2" cartridge)
ACC
-----Original Message-----
From: Kumar Deuja [mailto:kdeuja AT kdeuja DOT com]
Sent: Wednesday, August 14, 2002 3:11 AM
To: JimH AT FederalEdge DOT com
Cc: 'Kilesm'; veritas-bu AT mailman.eng.auburn DOT edu
Subject: RE: [Veritas-bu] StorageTek L180
Hi,
I was able to scan through every device file and found for files
that actually responded it has a tape drive, which are;
babu# mt -f /dev/rmt/15 status
/dev/rmt/15: no tape loaded or drive offline
babu# mt -f /dev/rmt/14 status
/dev/rmt/14: no tape loaded or drive offline
babu# mt -f /dev/rmt/13 status
/dev/rmt/13: no tape loaded or drive offline
babu# mt -f /dev/rmt/12 status
/dev/rmt/12: no tape loaded or drive offline
Then I added four drives using these files for no rewind files:
Drive0 -> /dev/rmt/15cbn
Drive1 -> /dev/rmt/14cbn
Drive2 -> /dev/rmt/13cbn
Drive3 -> /dev/rmt/12cbn
The drive now stays up but when I initiated a job then it goes down
as soon as the picker inserts the tape. It just can't seem to write to it.
Also, I ran the robtest and this is what I get.
Configured robots with local control supporting test utilities:
TLD(0) robotic path = /dev/sg/c2t0l0
Robot Selection
---------------
1) TLD 0
2) none/quit
Enter choice: 1
Robot selected: TLD(0) robotic path = /dev/sg/c2t0l0
Invoking robotic test utility:
/usr/openv/volmgr/bin/tldtest -r /dev/sg/c2t0l0 -d1 /dev/rmt/12cbn -d2
/dev/rmt/13cbn -d3 /dev/rmt/14cbn -d4 /dev/rmt/15cbn
Opening /dev/sg/c2t0l0
mode_sense failed, CHECK CONDITION
sense key = 0x2, asc = 0x4, ascq = 0x3, LU IS NOT READY, MANUAL INTERVENTION
REQ'D Enter tld commands (? returns help information)
Now, I have no clue why its failing, any suggestions or ideas would
be appreciated.
Best Regards,
=======================
Kumar Deuja
kdeuja AT kdeuja DOT com
page:ekdeuja AT kdeuja DOT com
=======================
On Tue, 13 Aug 2002, Jim Horalek (JimH AT FederalEdge DOT com) wrote:
> I think the L700 numbered drives drive with one on the bottom not on
> the top. The L180 may number the drives in reverse as well. Symptom is
> the wrong drive is loaded and ends up being downed. Then it tries
> again with the next drive. You can check this with robtest and then
> use "mt" status(assuming Unix) to check if the drive is loaded at the
> correct address.
>
> Jim
>
> -----Original Message-----
> From: veritas-bu-admin AT mailman.eng.auburn DOT edu
> [mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu]On Behalf Of Kilesm
> Sent: Tuesday, August 13, 2002 9:51 AM
> To: Kumar Deuja; veritas-bu AT mailman.eng.auburn DOT edu
> Subject: Re: [Veritas-bu] StorageTek L180
>
>
> Two issues that I can see
>
> First, you should be able to see all 6 drives, check
> your st.conf or sg.conf so that you have support for
> LUNs higher than 0.
>
> Secondly, from sgscan drives show up as 8,9,10,11cbn,
> and if in NetBackup they are defined as 0,1,2,3cbn
> then NetBackup will down these drives.
>
> Mike
> --- Kumar Deuja <kdeuja AT kdeuja DOT com> wrote:
> > Hi,
> >
> > I apologize sending this e-mail to you. I
> > saw your posting at
> > veritas-bu and you had encountered a similar problem
> > the one I am having
> > right now. I am using Vertias NetBackup V 3.4 with
> > Solaris 5.7. I have
> > StorageTek fiber switch connects to a scsi router
> > which is connected to
> > all the drives from L180 and the fiber switch is
> > also connected to fiber
> > channel interface to Sun450.i I used sgscan under
> > /usr/openv/volmgr/bin/sgscan and it seems to detect the hard drives
> > as well as the robot L180 and the drives. Actually the
> > L180 has 6 drives but
> > it only detects four drives. I tried adding the
> > drives and assigned the no
> > rewind files as follow:
> >
> > Drive1 /dev/rmt/0cbn
> > Drive2 /dev/rmt/1cbn
> > Drive3 /dev/rmt/2cbn
> > Drive4 /dev/rmt/3cbn
> >
> > When I use sgscan, this is the respond I
> > get.
> >
> > /dev/sg/c4t1l2: (/dev/rmt/8): "STK 9840"
> > /dev/sg/c4t3l0: (/dev/rmt/9): "STK 9840"
> > /dev/sg/c4t3l1: (/dev/rmt/10): "STK 9840"
> > /dev/sg/c4t3l2: (/dev/rmt/11): "STK 9840"
> >
> > As soon as I add these drives and start the
> > media manager daemon,
> > all drives comes down together, doesn't even stay up
> > for a second.
> > However, I can update the inventory using the media
> > manager without any
> > problem. If you have any suggestion or advice for
> > me, I would really
> > appreciate it. I am little new to this and I didn't expected to be
> > this complicated. Any help would be greatly appreciated.
> >
> > Once again, thank you so much for your time.
> >
> >
> > Best Regards,
> > =======================
> > Kumar Deuja
> > kdeuja AT kdeuja DOT com
> > page:ekdeuja AT kdeuja DOT com
> > =======================
> >
> > _______________________________________________
> > Veritas-bu maillist -
> > Veritas-bu AT mailman.eng.auburn DOT edu
> >
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
>
>
> __________________________________________________
> Do You Yahoo!?
> HotJobs - Search Thousands of New Jobs
> http://www.hotjobs.com _______________________________________________
> Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
>
> -------------------------------------------
> Introducing NetZero Long Distance
> Unlimited Long Distance only $29.95/ month!
> Sign Up Today! www.netzerolongdistance.com
>
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|