Ok, there has to be some special incantation here.

I have tried both forms of this command and have not bee successful.

How does one configure the environment, preferably on OpenVMS for
invoking SQLplus so the transport script will run?

I have also seen the sequence of operations on the "ON TARGET" form of
the procedure different on various sites. It is not clear whether the
RMAN CONVERT should be before or after the SQLplus execution of the
transfer script.

Any insight or guidance is appreciated.

Thanks,

Bill.

On Mar 29, 6:06*pm, BillPedersen <wmapeder... (AT) gmail (DOT) com> wrote:

Bill,

What is your target system? See note 413586.1 According to this
note, you must copy all of the resulting files to the new platform and
execute them on the new target system. Did I miss something or is
OpenVMS the target??? Make sure your symbols for SQLPlus are
properly defined as well as the logicals for ORACLE_HOME.

Error messages? DB Version? Source and Target OS/version?

On Mar 29, 7:03*pm, onedbguru <onedbg... (AT) yahoo (DOT) com> wrote:
Source DB: "Linux IA (32-bit)"
Target DB: "HP Open VMS"
RDBMS is 10gR2 on both systems
OpenVMS I64 V8.4
RHEL 4.8

If I follow the "procedure" for "convert" source then that is the
"simplest" as it does not require RMAN execution on OpenVMS.

Yes, copied the files to destination, in fact have placed them in the
expected directory on the VMS system.

But then when it comes to executing the transport script (which has
been edited for the target environment) then if I have Oracle enabled
"normally" it already has a database open so it starts to complain
that the there is already a database and can not complete the script.

If I have Oracle started but no DB open then it complains as well,
executing orauser.com without a SID. Then I end up with:

LEEDS$ sqlplus /nolog

SQL*Plus: Release 10.2.0.2.0 - Production on Tue Mar 29 19:33:26 2011

Copyright (c) 1982, 2005, Oracle. All Rights Reserved.

SQL> connect / as sysdba
ERROR:
ORA-12162: TNS:net service name is incorrectly specified

If I execute orauser against a known SID then it tries to do
everything again that known database instance.

Do I need to create a database with a different SID and then replace
the files there? That is not laid out in any of the documents I have
seen.

Getting to the notes, 413586.1, seems to be problematic for me. I
have tried but with the new environment Oracle has been setting up I
do not seem to be able to... No support right now, just a small
development/porting effort to get a product going and then...

Thoughts?

Thanks,

Bill.

Bill:

# Thoughts?

How about a straight up impdp using network link?

On Mar 29, 8:59*pm, John Hurley <hurleyjo... (AT) yahoo (DOT) com> wrote:
Well, being the newbie on this Oracle RDBMS probably because I was
unaware.

I will investigate.

Bill.

On Mar 29, 8:59*pm, John Hurley <hurleyjo... (AT) yahoo (DOT) com> wrote:
I guess the other thing is I did not see much/any discussion on it in
the 10g2R documentation set...

Found some notes online, will review.

From what I can see, I would suspect I need to have "target" database
in the destination - one which can be "blown away" and then it is a
matter of logging into both systems, setting up the link and then
starting the expdp and then the impdp processes on SQLplus with the
full=y option on the expdp.

Does one need to do the same sort of PL/SQL recompile on the target as
discussed with RMAN CONVERT process? I did not yet see any discussion
around that.

Thanks,

Bill.

Bill:

Yeah doing this requires the target database already created/running/
etc ... typically with tablespaces allocated/created already.

Using the network option the impdp gets the stuff from the source
database using a network link so you don't do an expdp ... just
impdp. Or you can expdp to a file ... then impdp from a file and
avoid network link.

I typically script it up with a bunch of steps like dropping the
receiving side schema owners first, creating/recreating the link and
directory objects ... then impdp using network link.

Shouldn't have to do "much" with recompiling stuff after ... but I
always have a procedure in each schema that recompiles stuff in it ...
and usually hit that at the end of any signficant refresh ( or
application schema change etc ).

On Mar 29, 8:26*pm, BillPedersen <wmapeder... (AT) gmail (DOT) com> wrote:
You don't have to, since things are automatically compiled as needed,
but you may want to, especially if you have complex dependencies (and
you want to find any problems beforehand, such as missed object
privileges). See http://www.oracle-base.com/articles/...emaObjects.php

I don't on my systems, but they aren't all that complex. They are
kind of strange for this type of operation, though, there are extra
operations I have to do. It all depends on the particular app what
the best thing to do is. I have to ignore a lot of "table doesn't
exist" types of errors on import which get resolved on recompilation,
and "user doesn't exist" errors which are informative, but useless.
You have to understand all errors. Having MOS access can really help
with this, especially when it is a new world. Everyone goes through
the same learning curve, and that's a basic repository for solutions.
Heck, I probably couldn't pass a job interview now, everything I've
forgotten is on the net.

jg
--
@home.com is bogus.
http://www.theregister.co.uk/2011/03...gence_gravity/

"BillPedersen" <wmapedersen (AT) gmail (DOT) com> wrote

On Mar 29, 8:59 pm, John Hurley <hurleyjo... (AT) yahoo (DOT) com> wrote:
Well, don't think you'll go very far with it on OpenVMS...

Have you read the announcements from Oracle, just last week ? No more
development on *all* Itanium platforms.
You have OpenVMS I64 V8.4 (me too, sadly)
Maybe you should reconsider this migration...

Syltrem

On Mar 30, 12:55*pm, "Syltrem" <syltremz... (AT) videotron (DOT) ca> wrote:
Here's another view: http://www.theregister.co.uk/2011/03...tanium_gambit/

But as someone who is on hp-ux Itanium, who remembers the 7.2 to 7.3
Oracle upgrade requiring a VMS upgrade, who remembers DEC forcing the
death of PDP in favor of VAX even while an 11/70 could handle higher I/
O loads than any VAX... I'm not real sympathetic, and at the same
time, having to agree with the anti-Larry sentiment I've been working
with for years.

At least we're not on Rdb: http://www.openvms.org/stories.php?s.../03/30/2407593

jg
--
@home.com is bogus.
If at first you don't succeed, please don't "help" the less
fortunate. http://www.signonsandiego.com/news/2...boat-accident/