Sun Blade 1500 Boot Warning Message 2

[bryan408]

do a search on google i found out this website very interesting As a newbie, this is a first new problem for y'all pros in here

Whenever i boot up a SUN BLADE 1500 with a new 120GB hardrive Seagate ATA Model: ST3120814 (not from Sun), i always got this warning message
Warning: /pci@1e,600000/ide@d/dad@0,0 (dad0):
corrupt label - wrong magic number
how can i get rid of this message?

This SUN BLADE 1500 runs on Solaris 8, OBP 4.17.1, 1GB mem, Serial # 63233863, Ethernet Address 0:3:ba:c4:df:47
and the original SEAGATE hardrive from SUN is ATA model ST3120026A

 

[tuka]

Run format as "root" and select the new disk (ST3120814).

It will ask if you want to label, select yes.

If you want, you can partition the new disk. Remember to
label the drive after partitioning.

exit from the format utility.

 

[bryan408]

Yes i already did label the disk at the CDE. The warning message can only be cleared at the console message but not at the OBP prompt

 

[bryan408]

come on cndcadams! i need some help

 

[kHz]

tuka is right in his answer. Your disk doesn't have a Solaris label.

 

[cndcadams]

bryan408;

What up? The above answers from tuka and khz are right on, both high quality engineers/Admins. If you do label the same drive and everytime you reboot, it comes up with corrupt label I would junk it. Basically it is telling you it can't hold a label. The message you see on bootup is a new message each time, this is not a message that can be marked as fixed like event logger.


If you are saying that everytime you put in a different 120GB drive it comes up with corrupt label it is most likely because the vendor they are getting the drive from has not run it in a solaris box and therefore you need to label it.

Thanks

CA

 

[bryan408]

Tuka, this is exactly what i did but i still had the warning message "wrong magic number" every time i boot up the system

Boot up the system and log in as "root"

# format <--'
Available disk selection:
0. c0t0d0 <ST3120814A cyl 57459 alt 2 hd 16 sec 255>
/pci@1e,600000 /ide@d/dad@ 0,0
Specify disk <enter its number> : 0 <--'
format> label <--'
Ready to label disk, continue? y <--'


format> par <--'
partition> label <--'
Ready to label disk, continue? y <--'
partition> q <--'
format> q <--'
# halt

Again i still can't get rid of that warning message. Should i reload the Solaris 8 though?
thank you very much for all replies. I really appreciate your help.
Regards,
Bryan

 

[tuka]

How many disk(s) do you have in this 1500?

From your format output I only see one disk. What are you booting from?

 

[bryan408]

tuka,
Originally, i have 2x120GB Seagate hardrives disk0 and disk1 but i took disk1 out and boot up from disk0 (c0t0d0)

if i leave 2 disks in there (still booting up from disk0), it will show 2 warning messages
Warning: /pci@1e,600000/ide@d/dad@0,0 (dad0):
corrupt label - wrong magic number

Warning: /pci@1e,600000/ide@d/dad@1,0 (dad1):
corrupt label - wrong magic number

regards,
Bryan

 

[tuka]

Does the O/S operate properly after boot?

what does the format/verify show?

Some of the reasons for the corrupt label and wrong magic number are:

1. something wrote non-label data to the block 0
2. the disk has failed in such a way as to corrupt block 0
3. data from block 0 was corrupted in transit.

Since both disks appear to suffer from the same problem the probability is that this problem was induced during the initial setup.
************************************************************Can you boot single from the cdrom media : " boot cdrom -s"

After booting to single user from the cdrom, run format and
select the new disk (ST3120814) and label the disk.

Then run verify and check the partitions, do they look okay?

Can you mount the new disk (mount to /mnt)

 

[cndcadams]

bryan408;

If you boot cdrom -s you still see the error correct?


Do you have any of these drives that are actually from Sun. If yes do they get the same error?

Also in the past I have seen where third party ide drives don't work very well with Sun/solaris. The problem is you can't format the ide devices (or could not a few years ago, not sure if this has changed). Check this out; info doc information;

**************
"format" option of format utility does not work for IDE drives. The reason for that is we can not do a low level format of an IDE disk. These disks come pre-formatted from the vendor. The servo, sector layout and defect management
information contained in the low level format is designed to last the life of the drive. This information and the way it is written cannot be duplicated in any way outside of the factory, thus we can not do a low level format.

On doing low level format we would get the following message for IDE disks "Cannot format this drive. Please use your Manufacturer supplied formatting utility."

Same applies to "defect" option too, where we will get the message "Controller does not support defect management
or disk supports automatic defect management.


*************

It is strange that you would be able to boot from it. The error is saying the label is corrupt so I am not sure how it can boot.

Run format choose type, then choose the correct label then label it and see if it clears up the problem. This however may blow away your OS.


If these are for a customer then you should contact the vendor you got them from and make sure they are testing these in Sun systems ( I doubt it).

Thanks

CA

 

[bryan408]

Does the O/S operate properly after boot? Yes

What does the format/verify show?

FORMAT MENU:
disk - select a disk
type - select (define) a disk type
partition - select (define) a partition table
current - describe the current disk
format - format and analyze the disk
repair - repair a defective sector
show - translate a disk address
label - write label to the disk
analyze - surface analysis
defect - defect list management
backup - search for backup labels
verify - read and display labels
save - save new disk/partition definitions
volname - set 8-character volume name
!<cmd> - execute <cmd>, then return
quit
format> verify

Primary label contents:

Volume name = < >
ascii name = <ST3120814A cyl 57459 alt 2 hd 16 sec 255>
pcyl = 57461
ncyl = 57459
acyl = 2
nhead = 16
nsect = 255
Part Tag Flag Cylinders Size Blocks
0 root wm 0 - 1003 1.95GB (1004/0/0) 4096320
1 swap wu 1004 - 1518 1.00GB (515/0/0) 2101200
2 backup wm 0 - 57458 111.79GB (57459/0/0) 234432720
3 unassigned wm 0 0 (0/0/0) 0
4 unassigned wm 0 0 (0/0/0) 0
5 unassigned wm 1519 - 2522 1.95GB (1004/0/0) 4096320
6 usr wm 2523 - 3526 1.95GB (1004/0/0) 4096320
7 unassigned wm 3527 - 8546 9.77GB (5020/0/0) 20481600

format> q

************can you boot single from the cdrom media: "boot cdrom -s" Yes and it still shows the warning message during the boot-up process

boot cdrom -s
.......
.......INIT: SINGLE USER MODE
# format

after running verify and partition they'd show the same output as above
#halt

after reboot the system, warning message are still showned

can you mount the new disk (mount to /mnt) Sorry i'm newbie i dun know how to mount a disk

regards
Bryan

 

[bryan408]

cndcadams,
you are right cndcadams. These drives are from the third party. If i use the originally drive from Sun, i don't have any problem.

actually i can boot up from boot cdrom -s but i can't format the disk and the message is Cannot format this drive. Please use your Manufacturer supplied formatting utility." Then i reload with a Solaris 8 (02/04) and i still have the same problem though

Anyway, there is no way to get rid of these warning message???
regards,
Bryan

 

[cndcadams]

bryan408;

So if you loaded the sun drive with solaris 8 you do not receive the error on bootup correct?

Also notice that the Sun drive has a different ST # then the ones you got from the vendor. Sun does not seem to support the other model you have. Not that I really think it is a model issue, but more of who the OEM actually built the drive for and how they set it up in manufacturing. Sun tends to require some tweeks for there ide stuff to work properly. If it were a scsi drive all you would need to do is reformat the drive which would add a sun label, update the firmware per sun requirement and you would be good to go.

Anyways I would have purchaseing contact the vendor and see if the vendor has sun labeled drives which they tested good in sun systems.

thanks

CA

 

[bryan408]

I ain't got no problem when loading solaris 8 with Sun drive. And yep, SUN drive does have a different ST# <ST3120026A> compared with a third-party drive <ST3120814>
Unfortunately, Sun doesn't sell these 120GB <ST3120026A> drives anymore and that is the reason why i have to get them from a different vendor.

 

[bryan408]

while running the sunvts test to check on the new drive (the one from third-party <ST3120814>), I have something really strange. The time it takes to pop up the test environment is pretty longer than normal. And also there are some warning messages on the console which i really don't understand at all

# uata : Warning : timeout : reset target chno = 0 targ = 0
# uata : Warning : timeout : reset bus chno = 0 targ = 0
# uata : Warning : Changing the mode of targ : 0 to ultra DMA mode : 4
# smt : smt0 : uf FDDI driver is not active. Initialization of this driver cannot be completed


best regards,
Bryan

 

[tuka]

If you want to peek at the new disk and see what is actually written as a label, run these commands.
This will be true if the boot disk is the new disk (ST3120814)
dd if=/dev/dsk/c0t0d0s2 of=tmp/diskdump.dat bs=512 count=1


***This will dump the first block of disk data into the file
/tmp/diskdump.dat.

**Now view the data using the octal dump utility. use the argument 0x0 to show the offset in hex starting from offset 0.

od -c /tmp/diskdump.dat 0x0

You will be able to see what is written on the disk

 

[cndcadams]

bryan408;

Hey if your up to trying something do this.

put the second drive (third party drive) back in. boot your system then run the below command to your second drive

dd if=/dev/zero of=/dev/rdsk/c0t1d0s0 count=1024

then relabel the drive, reboot and see if the error still exists on the second drive.


the dd command I pulled from another thread where someones ide drive became corrupt and they were not able to reload the system until someone had them run this command and it cleared up there issue. It was not an exact match to your issue but I for one am interested in seeing what would happen.

Thanks

CA

 

[bryan408]

hi Tuka,
this is what i got from that command

f=dev/dsk/c0t0d0s2 of=tmp/diskdump.dat bs=512 count=1
1+0 records in
1+0 records out
# od -c /tmp/diskdump.dat 0x0
0000000 S T 3 1 2 0 8 1 4 A c y l 5
0000010 7 4 5 9 a l t 2 h d 1 6
0000020 s e c 2 5 5 \0 \0 \0 \0 \0 \0 \0 \0
0000030 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0
*
0000080 \0 \0 \0 001 \0 \0 \0 \0 \0 \0 \0 \0 \0 \b \0 002
0000090 \0 \0 \0 003 \0 001 \0 005 \0 \0 \0 \0 \0 \0 \0 \0
00000a0 \0 \0 \0 \0 \0 \0 \0 004 \0 \0 \0 \0 \0 \0 \0 \0
00000b0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 ` \r 336 356
00000c0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0
*
00001a0 \0 \0 \0 \0 025 030 340 u \0 \0 \0 \0 \0 \0 \0 001
00001b0 340 s \0 002 \0 020 \0 377 \0 \0 \0 \0 \0 \0 \0 \0
00001c0 \0 > 201 @ \0 \0 003 354 \0 017 320 \0 \0 \0 \0
00001d0 \r 371 ( 320 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0
00001e0 \0 \0 \0 \0 \0 \0 005 357 \0 > 201 @ \0 \0 \t 333
00001f0 \0 > 201 @ \0 \0 \r 307 001 8 206 @ 332 276 9 8
0000200
#

cndcadams,
i already tried what you've told me to do and i still saw that warning message on both drives

best regards,
Bryan

 

[tuka]

Bryan,

That output looks okay.

Lets look at the whole disk label in hex and look for VTOC_SANE AND DKL_MAGIC .

Use the same file as earlier (/tmp/diskdump.dat)

od -x c0t0d0s2 0x0

(VTOC_SANE is always 0x600DDEEE at offset 0xbc)
(DKL_MAGIC is always 0xDABE at offset 0x1fc, jsut before the checksum.

You saw the error when booting "single" from cdrom. When booting single from cdrom is the error seen when only the original disk is installed?