Hardware

Documention for the Sunfire V210:

http://docs.oracle.com/cd/E19088-01/v210.srvr/index.html

It's a server.  Servers don't "do" monitors.

They need to be customized to use them and unless you have a Sun USB keyboard and monitor attached, there won't be any output to a screen.

No keyboard/mouse?  They'll default back to the Network Management port or else to the OBP output at the serial console port on the rear of the chassis (whichever it's been configured to use).

Documentation for the Ultra-1 :

http://docs.oracle.com/cd/E19127-01/ultra1.ws/index.html

Note that the older, first generation U-1's had three SBUS expansion slots and a HD-50 Narrow SCSI integrated to the systemboard.

The later generation U-1's (also refered to as the "Ultra-1E" systems had two SBUS expansion slots on the rear righthand side, a framebuffer (graphics) slot on the reat left and HD68 SCSI integrated to the systemboard.

Non-E Ultra-1's topped out at 170MHz CPU speed.  Ultra-1E systems could be as magically fast as 200MHz !!

Ultra-1's were discontinued in 1997.

Ultra-1E's were discontinued in 1998.

(Sunfire V210's were EOL'd around 2007)

That external disk won't work with the V210.  It's for the Ultra-1.

Good luck to you and happy reading.

Thank you so much for the helpful advice!

I have a Sun keyboard.  The mouse was in pieces so I will need to replace that.    I'm confused about the mouse and monitor part.  Can you clarify?  Do the monitor and the mouse both have to be Sun also?  In spite of my confusion on this point, I do understand that even with the proper keyboard/mouse/monitor, it may not give monitor output depending on the current configuration.

And, suppose I have the right keyboard/mouse/monitor but that this box is locked up very well.   As I don't have the password and don't have the cooperation of the original owner/company that surrendered it, are there options?   For example, in Microsoft environment, I could reset a password possibly, or pull a drive and get the data from the drive using an external enclosure.  On a cisco firewall, I can reset the password...etc

Can you give me a rough opinion on how likely it is that I will succeed here.  I can always purchase the Sun peripherals mouse/keyboard/monitor on ebay. But what about getting past the protected logins and built in security?

Using Sun mouse or monitor is not mandatory with this platform.

And for the password reset, the procedure can differ depending which file system and volume manager software are used. But basically, you need to boot the server with an CD-ROM, and DVD or with a Jumpstart server. Then you should mount the root partition, edit the /etc/shadow file and remove the encrypted password field on "root" line. And before to reboot, please unmount the partition.

Can I use an Ubuntu DVD to boot?  I tried to download Jumpstart and it's a 1.5MB .pkg file.  Not sure what to do with it...

If you have a Ubuntu distribution that supports SPARC processor, you can then try.

Jumpstart package ? 1.5 MB package size ? What is this file ?

Why not just booting with a Solaris media ?

I put the solaris disc in the cdrom and booted... This is the output I'm getting.  At the end of this where it says "Probing I/O buses" It's just sitting there at the very next line - for what seems like indefinite.  As I don't have a true command prompt to type at, can you tell me what is wrong?  What I'm missing?

SC Alert: Host System has Reset

Executing Power On Self Test
0>
0>@(#) Sun Fire[TM] V210/V240, Netra[TM] 210/240 POST 4.18.10 2006/03/03 14:19
       /export/delivery/delivery/4.18/4.18.10/post4.18.0/Fiesta/enxs/integrated  (root)
0>Copyright © 2006 Sun Microsystems, Inc. All rights reserved
  SUN PROPRIETARY/CONFIDENTIAL.
  Use is subject to license terms.
0>OBP->POST Call with %o0=00000800.01014000.
0>Diag level set to MAX.
0>Verbosity level set to NORMAL.
0>Start Selftest.....
0>CPUs present in system: 0 1
0>Test CPU(s)....Done
0>Interrupt Crosscall....Done
0>Init Memory....Done
0>PLL Reset....Done
0>Init Memory....Done
0>Test Memory....Done
0>Test CPU Caches....Done
0>Functional CPU Tests....Done
0>IO-Bridge Tests....Done
0>INFO:
0>      POST Passed all devices.
0>
0>POST: Return to OBP.

SC Alert: Host System has Reset

Configuring system memory & CPU(s)

Probing system devices
Probing memory
Probing I/O buses
ÿProbing system devices
Probing memory
Probing I/O buses

The diag-level is set as MAX. The checks can take some time to finish. Please wait and if you don't see any progress after 30 minutes, then you have a hardware issue.

It's now been 1.5 hours and it's still at this "Probing I/O buses".   So I guess I have a hardware error?  Any ideas on how to proceed?  I just need to get the data off.  I don't really need to resurrect the server, though it would be nice.  

Yep, I guess too that you have a serious HW issue.

Please open a service request to replace the system board or the riser card (I guess that the error is there). Or if you want just to get the data off, try to connect the appropriate disk in a server or in a storage array that support this disk.

It did finally get in.

The solution was to get to sc> and type break -y then sc> console, then wait.   It would then produce an "ok" prompt where I could specify to boot to cdrom -s. 

I'm in and trying to reset the root password so I can manage the data directly from the login with the root user.   But in reading online I was given to believe I need to reset this password on the boot drive (this is a server with mirrored drives).  I was given to believe further that I could identify this "boot drive" by entering into "Format" command.   But now as I'm in the command and believe I have the information I was after, I can't seem to get  out of the command without selecting a drive to format.  I don't want to format my drives.  I don't want to format either of them.   How can I exit this menu without formatting and overwriting my valuable data?

"q" for quit

or Ctrl-D

I found that by selecting a drive to format, it will only then display information and another submenu.  It's probably a good guess that this is what would happen.  But I had to take a very big risk in making this choice because loss of the information would have been unacceptable and it was clear that choosing a drive to format wouldn't "kick off" the process and begin the format.  

I'm able to quite the format dialogue from within the sub menu.   Thanks for the help.

Now that I'm "in" the server and can navigate some, I'm trying to reset the root password in the /etc/shadow file.   But I'm running into fragmented or inconsistent advice and documentation online about this process.  Clearly there is a lot of information about resetting the "root" password.   Much about doing this on a sun solaris server.  

The documentation and the advice is universal where it pertains to certain parts of the process.  It always recommends editing the shadow file "root" entry.  But that is where similarities end.  The path to the /etc/shadow file almost always includes the mount root; typically /a/etc/shadow.   But on my server there is not an /etc location under /a.   No /a/etc.  So I'm trying to find where the /etc directory is located.  Specifically the one on the "boot" drive (mirrored drives).

The shadow file that I've been into on my server contains an entry for root that looks like this, root:NP:6445::::::  

Advice on this has included, "NP means no password so there is none" (wasn't true in my case).   "Remove the NP, save, and then you can login"  (Didn't work).   "Remove the "NP" and the "6445", save, and then login (Didn't work).  And lastly, "That is the wrong shadow file" (seems to be a good possibility because it doesn't match the examples I see online where the root entry is followed by a hash string.)

So now my questions are; If it's important to know that the mirrored drives will play a role in trying to reset the password, then why doesn't the solaris documentation on resetting the root password refer to this?  Maybe it's not important and doesn't matter?  Or it does matter and the documentation does exist but I just haven't discovered it? 

And what is the definitive location for the /etc/shadow file that my server boots with?  How can I navigate there after booting to cdrom?  When I try to use all of the resources and documentation online, it includes the mount directory in the path.  My server does not have an /etc/shadow instance at the location when I use the mount directory in the path e.g. /a/etc/shadow.

# cd /a/etc

/a/etc: does not exist

#

Once I do actually find the right location, how exactly do I reset the password to blank so that it will result in my being able to access the root account?