Raars's Blog

We had a SUN server with Solaris 9 on it. We attached Momentum  disk raid , but OS did not see that RAID. Even after upgrading OS to Solaris 10 still no luck.

SCSI adapter was  Qlogic .

Here is a solution we did :

1) create /kernel/drv/qus.conf and add these lines:

bash-3.2# cat /kernel/drv/qus.conf

name=”qus” parent=”/pci@8,600000/pci@1″

unit-address=”4″

scsi-options=0x178 target0-sync-speed=0x27100 scsi-initiator-id=6;

2) modify /kernel/drv/sd.conf and add these lines ( though i think it is not necessary or we need add just one line )

name=”sd” parent=”qus”

target=4 lun=0;

name=”sd” parent=”qus”

target=5 lun=0;

name=”sd” parent=”qus”

target=0 lun=0;

3) go to OBP (OK prompt )

setenv auto-boot? false

reset-all

probe-scsi-all

check path to DISK RAID ( in this case it was /pci@8,600000/pci@1/scsi@4 )

cd /pci@8,600000/pci@1/scsi@4

.properties

check scsi initiator id ( usually by default it is 7 )

you will see scsi-initiator-id 00000007

we need change to 6 ( may be because of conflicts ??? )

you type :

6 ” scsi initiator id” integer-property

do reset-all

cd /pci@8,600000/pci@1/scsi@4

.properties

and make sure this time scsi-initiatorid is 6

4) boot -rv

and disk was there !!! YEEEESSSSSSSS!!!

P.S. Ofcourse physical path  to disk RAID will be different in your case

on previous  Solaris versions  you had to modify  two files :

/etc/default/login  should have:

CONSOLE=/dev/console

  /etc/ssh/sshd_config  should have

PermitRootLogin yes

On Solaris 11  you have two modify  one more file :

/etc/user_attr

By default it is like that :

root::::type=role
user1::::lock_after_retries=no;profiles=System Administrator;roles=root

After you modify it should be like this:

#root::::type=role
root::::
user1::::lock_after_retries=no;profiles=System Administrator;roles=root

here is what happened :

* We had  – HP Integrity rx2620  server with Linux Enterprise Workstation installed on it.

Server had a 3 internal disk and attached RAID storage . Monitor , keyboard and mouse were attached to server

* Application owner tried to reboot machine and then called us saying machine never came up

* We find out one of disks on external storage is dead (for clarity , there was not any mirroring or parity )

* Found out that , when machine “hangs” after initrd  message . Waited 15-20 minutes , tried ping from remote machine , no answer

* turned off and turned on machine again and opened “boot configuration ” menu .

opened EFI shell and ran conconfig (console configuration ). It showed that , VGA was as a secondary console . made it primary and rebooted machine .

This time i got all output to the monitor and system came to single-user mode because of failed disk (could not mount file system ) . Removed entry from /etc/fstab and then rebooted machine.

Everything came up fine (except of course , failed file system)

1. How to increase file system ?
   

chfs -a size=<new size in 512 blocks> filesystem

Example:

lsfs /tmp

Name Nodename Mount Pt VFS Size Options Auto Accounting

/dev/hd3 — /tmp jfs2 7340032 — yes no

chfs -a size=7345000 /tmp

1. Increase swap space
   

First find swap space:

lsps –a

Then

chps -s <additional LP”s> -a y <swap space>

Example:

chps –s 2 –a y hd6