Migration Install 4.2.1 to 4.3.3 [ASAP]

[Janghyegyung]

I tried to install by migration just a few minutes ago.
On the menu, default setting displays hdisk3 for installed disk, but my systems's rootvg is on hdisk0. Strange thing is the location code is right for rootvg, hdisk0.

I cannot make decision how I can handle this situation. What's the right information for rootvg, disk name or location code?

Give me any advice. This server is on the list SLA and this time is the second. I'm so stressed and embarassed. Please..

 

[aixqueen]

My opinion it is the location code

lspv to get the disks in rootvg
then lsdev -Cc disk
and lsdev -Cc adapter
lsdev -Cs scsi

And whatever the location code is ...is where the disk physically is.....disks can get out of wack number wise because of ghost disks, (disks that were failed and replaced without removing first) So when it defines a disk it looks at the location code....

You have a backup of everything, yes? so if it did make a mistake you would ok....

I would go with location code.
----------------------------
From the IBM docs below

Installation Setup.....Change the Destination Disk

The Change Disks Where You Want to Install screen allows you to change the
hard disk where BOS will be installed. The location codes of the hard disks are
displayed in the Location Code column of the Change Disks Where You Want
to Install screen. The format for the location code for a direct-attached
disk is: AA-BB where AA is 00 (zero) and BB is the slot number for the
hard disk. The format for the location codes for all other hard
disks is described in "Location Codes," in POWERstation and
POWERserver Common Diagnostics Information Manual.

Note: You may wish to keep a record of the location code for the destination
disk. In the future, you can use this location code to identify which disk
contains the root volume group in order to do system maintenance.

Additional 4.3.3 stuff

http://publib.boulder.ibm.com/doc_link/en_US/a_doc_lib/aixins/aixinsgd/install_bos.htm

http://techsupport.services.ibm.com...mode=9&documents=091623329313242&database=aix

http://techsupport.services.ibm.com...mode=9&documents=990605188914736&database=aix

 

[hasp2]

It is a known issue that the disk #'s can change durring a migration install. ALWWAYS us the location code and you will be fine!.

 

[Janghyegyung]

Thanks for your response.
I kept going on, and it was ok.

Could you give me an answer for another?

During the patch, to end user, how will be the service? For instance, an user's trying telnetting or using DBMS? We say the time for patch or updating some filesets is no-service, or normal-service time?

 

[aixqueen]

I don't understand.......but if you are saying can the system be up when you apply maintenance patches............. my opinion is no...........

The system should to be without users.........except root......I even shut down the databases...etc.... and issue a nologin to keep users out.

Any patch with a bos in front of it...usually requires a reboot after applying and the
patch.......system should be as quiesed as possible....(my opinion)

I remember a user that applied bos.rte.tty while users were on the system, and the entire system gacked and he had to do a restore from mksysb....risky to do while users are up...It depends on the patch of course...but anything with bos at the beginning of the patch is even riskier....you have to reboot anyway...easier to get them off, do the patches and then reboot...and life is good.....

 

[Yegolev]

Assuming you are going from 4.3.3.0 base to the 09 patch set, I have seen a bug where paging is disabled after I upgraded to 4.3.3.0 base, but the bug is fixed somewhere before 09. You have probably already gone through with it, but just in case you encounter a hd6 that you cannot enable, just make a temporary paginglv and patch to 09.