[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference smurf::ase

Title:ase
Moderator:SMURF::GROSSO
Created:Thu Jul 29 1993
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2114
Total number of notes:7347

2070.0. "customer's upgrade woes" by NNTPD::"nieldd@mail.dec.com" (dave nield) Mon May 19 1997 14:07

Hi Folks,
I just got this from a customer, anyone care to comment:

Dave,

Thanks for your help. Here is a brief description on what happened during the
latest attempt:

1) 2 x 8200 running DUNIX V3.2-D & ASE V1.3
2) Only one shared bus (SCSI No 4)
3) Steps: tried simultaneous upgrade to save time
Removed ASE V1.3 and saved /usr/var/ase/config/asecdb on both machines
installupdate to V4.0
installupdate to V4.0B
installed latest patch which provide dupatch facilities and a few other things
Both machines appear to be running DUNIX V4.0B standalone okay.
4) Now to install TCR V1.4
On one system:

setld -l TCR140/kit
choose 3 - Mandatory subsets and Ref Pages
answer y to use the old ASE database found
answer n to edit the kernel config file
The ASE I/O bus renumbering tool has been invoked.
The messages output as follows:
It appears there are differences in the SCSI configuration since the last time
this tool (ase_fix_config) was run. Please re-assign the correct names to the
scsi controllers.
No previous assignments have been saved.
Select the controllers that define the shared ASE I/O buses.

I was confused with the above messages and could not work out the correct
answers at that time and some mistakes were made. After choosing the bus 4
with name scsi4, then I chose the option f) default and quit without making
any changes.
I should have chosen x after seeing scsi4 scsi4
When I examined the /usr/sys/conf/HOSTNAME I found there were no device
entries under scsi4 at all. Rather horrified !
Not knowing to trust the kernel built - I abandoned the upgrade. Since I has
saved a copy of the OS, I tried to delete TCR140 but it declined saying
clu_install not found. A reinstallation of TCR140 was also declined because of
the same reason. I returned the systems back to UNIX V3.2-D and ASEV1.3
Questions:
a) Are the simultaneous 2 upgrades outlined above possible/allowed?
b) What do the messages "It appears there are differences..." and "No previous
assignments..." mean ?
c) What is clu_install ?

Pak Ng - Birmingham University.

[Posted by WWW Notes gateway]
T.RTitleUserPersonal
Name
DateLines
2070.1UNXA::PARKERMon May 19 1997 20:459
Just for the record, this is the same customer whose inquiry I posted in 2051. 
Unfortunately, I didn't receive any replies to my posting.  Dave, I hope you
have better luck here.

The customer is extremely aggravated and frustrated.  FYI - I've helped this
customer out in the past - now he comes to me everytime he needs help and the
CSC isn't helpful.  :(  Hopefully, somebody can come through with answers.

Dave
2070.2possibly another patch neededDYOSW5::WILDERDoes virtual reality get swapped?Tue May 20 1997 00:299
    This will not solve the problem you mentioned, but may solve a problem
    you may not know you have.
    
    I see that you hace 2 8200 systems with now TCR 1.4. If you are running
    OPS, you MUST get an additional patch from the CSC concerning a problem
    with remote drd. This patch is NOT in the current dupatch kit. If you
    are NOT running OPS, ignore this reply.
    
    /jim
2070.38200 only?NNTPD::"carterrg@mail.dec.com"Richard CarterTue May 20 1997 19:2611
Hi,

Is this OPS patch 8200 only?

I have a customer with 2*AS8200 with OPS+TCR1.4+V4.0B

If I need a patch, which number is it?

Regards,
Richard. 
[Posted by WWW Notes gateway]
2070.4this is a TCR patchDYOSW5::WILDERDoes virtual reality get swapped?Tue May 20 1997 20:2113
    This is a TCR patch. I believe there is an official patch thru the CSC
    fot TCR V1.4a. What is needed is a bss_rm.o file and then the kernel
    rebuilt. This .o file must be replaced on ALL 8200 systems that are
    running OPS. Contact the CSC. They should have the patch for TCR V1.4
    and 1.4a.
    
    What happens is that if you do a remote drd read/write, the i/o will
    never complete and the system will hang. This applies to all 8200, and
    any 8400 that does not have a cpu in slot 0 (there is no slot 0 on an
    8200).
    
    /jim
    
2070.5there are no official patches for 1.4/1.4aRHETT::MOORETue May 20 1997 20:308
    re .4 --
    
    We in the CSC have received no official patches for TCR 1.4 or 1.4a.
    Perhaps they are in the release pipeline.  
    
    Martin Moore
    Digital UNIX Support Group
    Atlanta CSC
2070.6All on the same page?NNTPD::"cherkus@buff.zk3.dec.com"Dave CherkusTue May 20 1997 21:3913
We are talking about the problem described in note 1578.

This is a TCR PSS problem, not an ASE problem!

The suggestion there was to use dbx to patch bss_rm_iodone_bind
to 0.

Doing a COMET search I found a memo that seemed to imply that the
dbx patch was the sanctioned fix, but to be honest I have a hard
time understanding the beaurocratic language in those memos.

Dave
[Posted by WWW Notes gateway]