• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1299
  • Last Modified:

Sun E4800 devalias boot problems

you know how it goes...box goes down, your life is hell until its up...

so with that being said, the E4800 was sitting at the ok prompt, It was due to a kernel panic.  I synced and reset the box and then at the "ok" prompt(auto-boot false, corp standard) I issued the standard "boot disk" and i kept getting a "ERROR: cannot open boot-device".  At that point I did a probe-scsi-all and my two 36gb fujitsu disks showed up.
Target 0
  Unit 0   Disk     FUJITSU MAN3367M SUN36G 1804
Target 1
  Unit 0   Disk     FUJITSU MAN3367M SUN36G 1804
Target 2
  Unit 0   Processor     Sun     D240            1.09
Target 3
  Unit 0   Processor     Sun     D240            1.09
Target 4
  Unit 0   Removable Tape     HP      C5683A          C005
Target 6
  Unit 0   Removable Read Only device    TOSHIBA DVD-ROM SD-M14011009

From there I did a devalias and this showed up....


{10} ok devalias
Alias                 Device Path
xdisk                 /ssm@0,0/pci@1c,700000/lpfc@3/sd@0,0
xnet                  /ssm@0,0/pci@19,700000/pci@1/SUNW,hme@0,1
diskc                 /ssm@0,0/pci@1c,700000/pci@1/SUNW,isptwo@4/sd@0,0
netc                  /ssm@0,0/pci@1c,700000/pci@1/SUNW,hme
cdrom                 /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@6,0:f
disk                  /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@0,0
disk0                 /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@0,0
disk1                 /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@1,0
disk2                 /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@2,0
disk3                 /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@3,0
disk4                 /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@4,0
disk5                 /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo@4/sd@5,0
def-cn                /sgcn
diskifp               /ssm@0,0/pci@19,700000/SUNW,ifp@1/ssd@0,0
diskglm               /ssm@0,0/pci@18,700000/scsi@2/disk@0,0
scsi                  /ssm@0,0/pci@18,700000/pci@1/SUNW,isptwo
net                   /ssm@0,0/pci@18,700000/pci@1/SUNW,hme

I tried to boot from disk, disk0 and disk1, but still got the same error.  Then I did a "printenv" and saw something interesting...

nvramrc=devalias vx-c2t1d0 /ssm@0,0/pci@19,700000/pci@1/SUNW,isptwo@4/sd@1,0:a

The root disks are mirrored with veritas(FYI), so a co-worker suggested i try to boot the exact path of the vx-c2t1d0 alias...and to my suprise, it worked?

e.g. <OK> boot /ssm@0,0/pci@19,700000/pci@1/SUNW,isptwo@4/sd@1,0:a

so I am trying to figure out what the ':a' signifies at the end of the device path, the dev alias look spot on except there missing the ':a' at the end of there, and they would not boot the system.  but booting to this exact device path with that ':a' brougt the box up.....what gives???? any ideas????
  • 2
2 Solutions
We stupid humans use the notation of s0, s1, s2, and so on to designate the partitions.  The OBP and OS (at a low level) as you see above use the notation :a (first partition), :b (second partition), and so to designate the partitions.  Ultimately, the c#t#d#s# files we all know and love (and understand) are symbolic links back to the ugly sd@#,#:<letter> files.

This more then likely means your primary boot disk has failed.  You need to run vxstat to ascertain the state of all your VxVM devices.  Even though your root disks are mirrored the OBP doesn't know that.  It still hits one disk intially as part of the boot cycle.  Then /etc/system is read and the VxVM device drives are force loaded which makes the other mirror disk start to be used.
I suggest you should use 'boot', not 'boot disk', first command will select default boot device, which is probably your vx mirror alias.
jedblackAuthor Commented:
thanks Nukfor, I had a feeling the ':a' was probably slice notations....

I told the sun tech the same thing about booting with regards to the vx mirrors.  He kept telling me that i should be booting to my vxmirror, and i kept saying that we are at the hardware level and not volume managment is running and i just need to boot from one of the disks on the internal array.  He was the one tha t finally suggested booting from that "nvramrc" variable that showed up in the 'printenv'

I'm still confused as to how the hell this box was booted in the past if the engineer was not sure of the exact device path to the boot disk?  I mean if you look at the 'devalias', they are all pointing to pci@18, I booted the box from pci@19 from the printenv nvramrc variable (nvramrc=devalias vx-c2t1d0 /ssm@0,0/pci@19,700000/pci@1/SUNW,isptwo@4/sd@1,0:a), is pci@19 just another scsi path to the same disks/devices listed on pci@18 in the 'devalias'?

also one last thing....what the heck does (nvramrc=devalias vx-c2t1d0 /ssm@0,0/pci@19,700000/pci@1/SUNW,isptwo@4/sd@1,0:a) mean?  I'm still sorta green with the sun hardware aspect, been doing os engineer work for awhile...but we have infrustructure engineers that handle the hardware setup/maintainence portion of the workload.... dont ya just love it when the workload micro-distributed(*sigh....)

thanks guys/gals

is pci@19 just another scsi path to the same disks/devices listed on pci@18 in the 'devalias'?
not always true, it's platform and your connection topology dependant
what is true - its another PCI card in another i/o board
to check all your scsi devices run 'probe-scsi-all' and refer to its output, not to 'devalias'

printenv boot-device

this variable defines default boot device when you type 'boot' without parameters or when auto-boot is performed.
Probably thats why this box was booted with a single 'boot' command. Or may be admin always was using 'boot vx-c2t1d0', who knows...
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Featured Post

Free Tool: IP Lookup

Get more info about an IP address or domain name, such as organization, abuse contacts and geolocation.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now