I am a new graduate student at UMKC-Physics. I have been put in
charge of getting a lab back up and running that has not ran in
quit some time. We have a Omicron sun4m running Solaris controlling
our STM. Luckily It had turned on, however it seems to have forgotten
its boot device, has hostid errors and doesn't remember its time
if shut off and turned back on. I believe this is because of a bad
NVRAM but would like to be damn sure before proceeding as if it
is. I believe this bad NVRAM is causing the workstation to not know
what its hostid is and therefore when the hostid is compared to
that listed in the license.dat file I get errors that keep me from
using my omicron software with the full privileges I have.
On 2018-06-13, cxb5h7@gmail.com <cxb5h7@gmail.com> wrote:
I am a new graduate student at UMKC-Physics. I have been put in
charge of getting a lab back up and running that has not ran in
quit some time. We have a Omicron sun4m running Solaris controlling
our STM. Luckily It had turned on, however it seems to have forgotten
its boot device, has hostid errors and doesn't remember its time
if shut off and turned back on. I believe this is because of a bad
NVRAM but would like to be damn sure before proceeding as if it
is. I believe this bad NVRAM is causing the workstation to not know
what its hostid is and therefore when the hostid is compared to
that listed in the license.dat file I get errors that keep me from
using my omicron software with the full privileges I have.
The general issues you describe certainly sound like a dead NVRAM,
can't comment on the licensing issues with your specific software
but tying to HostID seems quite plausible to me. Without any
specific issue I would say follow the FAQ - it's too long ago that
this cropped up for me to have anything worth adding except if you
have a specific issue.
The problem I do anticipate is recovering the original HostID,
which sounds like is going to be significant here. Ideally you
will have some documentation to refer to, for example I routinely
print off a dmesg after an OS install or major reconfiguration in
part for issues just like this. Sounds like they might be hard to
come by but may be worth asking around for.
Alternatively have a good look at the machine, inside and out: if
it has been used at all in the last 10-15 years it is likely this
is not the first time it has needed replacing and the previous chap
may have done you a favour and noted the HostID and/or MAC address
somewhere - I know I generally put them on a sticker either on the
PSU or the lid of the case for future reference - documents can
easily get lost after all.
The other place I'd look would the BOOTP (or more likely DHCP)
server for the LAN: even if the machine generally booted off the
local disk with a static IP address it isn't too unusual to find
an entry for the system there even if only for netbooting on OS
installation, especially likely if there is no optical drive. If
you can track down the MAC address for the system from there the
initial byte of the HostID is determined by the type of the system
and the remaining three bytes are the last three bytes of the MAC
address. Again, it's quite possible the DHCP server has inself
been replaced over the years but those settings are often copied
over without even considering whether the machine in question is
in active use, just to avoid the risk of breaking something.
--
Andrew Smallshaw
andrews@sdf.org
I have read through the FAQ and think I have a better understanding of things.
1.) get the corresponding chip and replace (model dependent )
2.) have the default values set (excluding the IDPROM info ) .
3.) set idprom with the procedure describe in the FAQ.
"General sun4c, sun4m, sun4d, sun4u IDPROM Programming" or do I use the
"A Quick-and-Dirty Guide to Restoring the NVRAM of a sun4c/m/u machine" ?
3.a.1) I will need to set the hostid, where the fist byte is set
specific to the machine. (Its here I get confused as I am unaware of
my model or how to find it. all I know is sun4m, sparc )
3.a.2) set last 3 bytes, which as you said can be found from the MAC id. (You listed a few ways to find this. The FAQ mentions a way to reconstruct the HOST ID and Ethernet id ..
5.b) How do I set the boot info as this is also stored by the NVRAM?
I am a new graduate student at UMKC-Physics. I have been put in charge
of getting a lab back up and running that has not ran in quit some time.
We have a Omicron sun4m running Solaris controlling our STM. Luckily It
had turned on, however it seems to have forgotten its boot device, has
hostid errors and doesn't remember its time if shut off and turned back
on. I believe this is because of a bad NVRAM but would like to be damn
sure before proceeding as if it is. I believe this bad NVRAM is causing
the workstation to not know what its hostid is and therefore when the
hostid is compared to that listed in the license.dat file I get errors
that keep me from using my omicron software with the full privileges I
have.
I have found this link, which seem very helpfully and pointed me here: http://www.menet.umn.edu/~bob/FAQ/sun-nvram-hostid.faq
I have little to know knowledge/experience with this computer, just
one week worth of trying to diagnose it. And my UNIX knowledge is
limited at best. I
I was hoping I might be able to get in contact with someone that might
be capable of walking me through a procedure to verify that this is
indeed the issue, and how to address it. As the last thing I wanna do I render this workstation UNBOOTABLE or mess up anything.
You can post here or email me directly at : cxb5h7@mail.umkc.edu
I would attach pictures of the errors I can find/teas out but I don't
see an attach button, please let me know how to do so or inquire for
them.
Abridged errors:
Forgotten boot device (disk)
Incorrect configuration checksum;
IDPROM contents are invalid
invalid format code in IDprom.
WARNING:TOD clock not initialized
LICENSE MANAGER: Invalid license key (inconsistent encryption code) (-8, 130) license file can not be checked out (FLEXlm error: -15)
inconsistent incription code for idl_rt
bad hostid format
hostid: hostid string returned by sysinfo not numaric ""
these are not all the errors but the ones that I fell or know to date
that support a bad NVRAM hypothesis.
I would also like to know if replacing the chip will effect my license
and if there is anything i can do do get it back up and running
The question is whether Oracle maintains the data to continue
this (replacement NVRAM) service. And whether they will do so.
I am wondering if/why/what is causing the workstation to load as sun and not Solaris.
UPDATE:
I have found documentation stating what I believe to be my original Hostid and Ethernet id.
Host-id: 808bc7f5
Ethernet id: 0:80:42:10:57:f5
From the first byte of the host ID and the FAQ, I surmise that the
system model is one of the following
SPARCstation Classic, LX, 4, 5, SS1000, Voyager, Ultra 1
But I don't understand which one it is? Or why both of these dot match
the conventional way of doing things.
Also the data sheet (customer sheet) where I found the hostid and the Ethernet id also lists:
SN#
Operating System: Solaris V2.5
Driver type: omicron v2.3
some setup info
When I use the command "uname -a" I get the following back
" SunOS scala 5.5 Generic_103093-08 Sun4m sparc sun4m"
The workstation information also says
"SunOS 5.5 Gen..."
And the start of the workstation it loads Sunos
I am wondering if/why/what is causing the workstation to load as sun
and not Solaris.
Also, I found this line in the FAQ
"The hostid on Solaris 2.5 x86
Intel processor machines don't have an IDPROM. Sun uses a different
mechanism to generate the hostid. When the operating system is initially installed a pseudo-random hostid is generated. It appears that this pseudo-randomly generated hostid will always be between 1 and 3b9aca00. The hostid is based on eight bytes of serialisation information in the kernel module /kernel/misc/sysinit. This is in contrast to the situation on SPARC machines where the hostid is based on the IDPROM.
/kernel/misc/sysinit contains code which initialises the variable hw_serial in the kernel based on the serialisation information. On both SPARC and x86 versions of Solaris 2.5, hw_serial stores the hostid as a decimal C string.
Other than the eight bytes of serialisation information the /kernel/misc/sysinit files do not differ between machines. Four of the serialisation bytes depend upon the other four bytes, so the hostid is somewhat tamper resistant. If the serialisation information is tampered with carelessly or the sysinit module fails to load for some other reason, the hostid of the machine will be 0. A little more obfuscation is done in the code, i.e. hw_serial is not referenced directly in the module, but
indirectly via the pointer _hs1107."
What I find confusing is whether or not to treat this workstation as a
sun, sparc, or solaris.
Furthermore, the line about the host id being zero is directly
applicable to what is happening.
the work station info yields it as a
zero and the licnse.dat file dose so to. I think I got hostid to say it
one time but now all i get is the aforementioned errors.
I have no idea the ramifications of this new info or how to proceed.
I have found what I believe to be a new NVRAM for the sun4m, but would
like to hold off on getting it till I am certain how to proceed.
I could not find a "/kernel/misc/sysinit " file
I do not know where to find a FLEXlm.dat license file
My thoughts: I assume that the pc is defaulting to the sun operation
system, the reason is unknown but it either has to do with the NVRAM or telling it "boot disk" .
You say "Omicron sun4m" -- is this a computer which was made by
Sun, and Omicron supplied it loaded with licensed software? --------------------------------------------------------------------------------------------Yes
I believe so
@Don , thanks for your response, Looks like it will be helpful, but at
the moment I do not fully understand what you are saying I will need to
think on it and formulate some questions.
There Is just so much new information involved with trying to get this
up and running figuring out what information is good and bad, much less keeping it all strait and understanding it is a bit difficult. I thank
you all for your help and patience on this matter.
If any of you are available to call at your convince , Being able to
verbally discus or bounce back ideas would be very helpful as this form
has been really the only and best source of information to me.
I have found documentation stating what I believe to be my original Hostid and Ethernet id.
Host-id: 808bc7f5
Ethernet id: 0:80:42:10:57:f5
From the first byte of the host ID and the FAQ, I surmise that the system model is one of the following
"SunOS 5.5 Gen..."=20
And the start of the workstation it loads Sunos
I am wondering if/why/what is causing the workstation to load as sun
and not Solaris.
"The hostid on Solaris 2.5 x86
What I find confusing is whether or not to treat this workstation as a sun,
sparc, or solaris.
Furthermore, the line about the host id being zero is directly
applicable to what is happening.
I do not know where to find a FLEXlm.dat license file
My thoughts: I assume that the pc is defaulting to the sun operation system= >, the reason is unknown but it either has to do with the NVRAM or telling i= >t "boot disk"
the keyboard/mouse/framebuffer/monitor are all unique
different technology that lost to the steam-roller of cheapo x86 PCs
taking over everything.
Yes, because your NVRAM battery is dead. You'll also find the Ethernet
MAC address is all zeros, and the machine can't talk on a LAN.
I wanted to give an update for to whom it may concern.
I have got my workstation up and running.
It was ended bad NVRAM-battery.
I had my department engineer take the Workstation apart and he found 2
NVRAM. I believe he said on was on the video board and the other on the mother board.
He has had experience with replacing NVRAMs in the past.
He clames that sometime despite being lucky enough to get you hands on a
new NVRAM it dosen't work.
So what he did was find where the battery was
with in the NVRAM using a magnet, took a magnifying glass and dermal
drill to cut open the top of the NVRAM to gain access to the battery. He
then wired up external coin bateries for both NVRAM. He did this in the
even that they leak they don't ruin the NVRAM.
We then put it all back together, Reprogrammed using the "## # mkp"
commands followed by "0 f 0 do i idprom@ xor loop f mkp" then "reset"
at first this put my WS in an infinite loop trying to boot from the
Ethernet to which it is not hooked up. But i learned that "stop + a"
gets me to the "ok" prompt. At which point "boot disk" worked with no
issue.
But, I would like to get the TOD set right. I still have issues with
this. I'm trying "date 081311452118" as the format should be date MMddhhmm[[cc]yy] where cc was to be century minus one. however I can
never get all the bytes to mach what i enter. For example. I may get the month and time right , but the day and year will be way off. If tried
many combos of this command and could not get it to work. The only way i could get it to a date 20xx was with ccyy=2418 but that yielded 2010.
which made my day and i believe time wrong. I have no idea whats going
on as there is no clear correlation between what i enter and what i get.
some times i enter 08 = MM and get July . If any one know how to solve
this last issue I would be grateful, well more grateful than I already
am. Thank you all for your help. Without it my lab would be nothing more
than scrap metal
Is it possible that you have a version old enough so it has not
been made Y2K compliant? There were patches as Y2K approached. The
version I have give the following in Solaris 2.6:
# sum /bin/date
9830 16 /bin/date
or:
# md5 /bin/date
MD5 (/bin/date) = 1f79e886fe37b3cce672c318788848b8
If your numbers are different perhaps you have a non Y2K compliant
version of the date command.
I could send you mine, if you want to accept a binary and try
it.
Good Luck,
DoN.
I entered "date 120305252426" to get aug 16 10:00 am 2018
Is it possible that you have a version old enough so it has not
been made Y2K compliant? There were patches as Y2K approached. The
version I have give the following in Solaris 2.6:
I got it to work , but oddly so. I entered "date 120305252426" to get
aug 16 10:00 am 2018
from what prompt do i use the eeprom command. it did not work in shell
eepromeeprom: Command not found.
eeprom testarea=0eeprom: cannot open /dev/openprom: Permission denied
Is it possible that you have a version old enough so it has not
been made Y2K compliant? There were patches as Y2K approached. The
version I have give the following in Solaris 2.6:
# sum /bin/date
9830 16 /bin/date
or:
# md5 /bin/date
MD5 (/bin/date) = 1f79e886fe37b3cce672c318788848b8
If your numbers are different perhaps you have a non Y2K compliant
version of the date command.
I could send you mine, if you want to accept a binary and try
it.
Good Luck,
DoN.
I got it to work , but oddly so. I entered "date 120305252426" to get aug 16 10:00 am 2018
from what prompt do i use the eeprom command. it did not work in shell
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 285 |
Nodes: | 16 (2 / 14) |
Uptime: | 70:30:41 |
Calls: | 6,488 |
Calls today: | 1 |
Files: | 12,096 |
Messages: | 5,275,488 |