Re: Req help: Oracle 8.0.5 Server Edition and RedHat Linux 6.1, creating a new SID

Hi Bill,
  well, you asked a question which would require quite a bit   of time to explain. I will do my best to answer quickly in   5-10 minutes so i can get back to my job - administering   Oracle development environment in my company. But before i   do that i strongly recommend to purchase a book by Michael Ault   "Oracle 8i Administration and Management" (and don't waste your   money on any of that Osborne/Oracle garbige). That book covers   mostly Unix env and should be sufficient for Linux. He has a book   for NT (I think it's called "Oracle on NT Black Book").   If you can't get it, the second best are books by Donald Burleson,   but i would go with Michael Ault first.

  In Oracle world, instance and database sometimes are two different   meanings and sometimes is the same thing - depending on the   situation and the subject,mostly because each user essentially   can have it's own database, refered to as schema. Therefore, when   someone says instance, he/she means the set of processes and global   memory, shared by all "databases/schemas" within that instance.   When you as a SYSDBA, create a new user and grant CONNECT and   RESOURCE roles to it along with some quota or unlimited quota   on some tablespace, that user can create objects, which will be   "stored" in that user's schema and noone except SYSDBA can change   them or even put any data in it. I mentioned tablespace. It's a   logical repository for Oracle objects. Different users can share   the same tablespaces, but if you would like to separate them for   maintenance purpose you should create separate tablespaces for   separate schemas, because when/if you need to do online(hot)   backups of those shared tablespaces noone would be able to change   any data in them during that time.
  Tablespaces consists of datafiles, which is Oracles's physical storage   plan carefully how you will spread them across the instance - this is   where your performance will come (of course, your SQL comes first,   but we are not touching it here).
  8.0.5 EE comes with Enterprise Manager, and it's not too bad. I   would strongly suggest to learn Oracle administration from the   command line if you are planning on making a carrier in that field.   Mainly because if you ever need to do any type of support remotely   you probably won't have any graphical interface and if don't   become intimately familiar with data dictionary and dba and v$ views   you will be in troubles. Once you are comfortable with those things,   you can use some GUI to speed up your work and save time.

In article <8dvsnh$nbo$1_at_nnrp1.deja.com>,   bill_moss_at_my-deja.com wrote:
> Hi Alex,
>
> I saw your Deja post on creating an Oracle DB on
> Linux. I recently installed Oracle 8.0.5 on Linux
> and was wondering about things like creating new
> instances, tablespaces, users, tables, etc. I
> sucsessfully set up the test DB and installed
> some tables from the O'Reilly SQL*Plus book - but
> did it in Scott/Tiger's schema - just was
> wondering how to do it for a new
> user/schema/instance. Can you recommend a
> resource to learn these basics (from the command
> line - i don't think my 8.0.5 linux ee comes with
> the enterprise manager) or do you know of any
> FAQ's, scripts, etc. I'll bet there's loads of
> people who could use this info! This is a good
> start but more on the other things would be
> helpful. Thanks in advance,
> Bill Moss
> bill.moss_at_hotmail.com
> take the . out of bill.moss
>
> In article <8dnbmd$q3l$1_at_nnrp1.deja.com>,
> asokol99_at_my-deja.com wrote:
> > Bruce:
> > in Korn (ksh)/Borne (sh) shells you use the
 following:
> > $ ORACLE_SID=sid1; export ORACLE_SID
> >
> > in C shell (csh) you use the following:
> > % setenv ORACLE_SID sid1
> >
> > since i usually use csh, i have an alias
 which i created in
> > .cshrc file:
> > alias sid "setenv ORACLE_SID \!*"
> >
> > so, when i need to switch to a new database
 instance i just use:
> > % sid newname
> >
> > RH uses bash as default shell ("Borne
 again") which behaves like
> > regular Borne shell in terms of environment
 variables.
> >
> > Alex
> >
> > In article <sfu1fbqkl6e28_at_corp.supernews.com>,
> > "Bruce" <bey_at_3eg.com> wrote:
> > > Alex,
> > >
> > > thanks for your input, I will did it your way
 before, I think my
 problem is
> > > using the right shell command for switching
 the instances. Do I use
 set
> > > oracle_sid=xxxx in the sh environment?
> > >
> > > Bruce
> > > asokol99_at_my-deja.com wrote in message <8dfnh2
 $akj$1_at_nnrp1.deja.com>...
> > > >Here are the steps to create a new database
 on Linux:
> > > >
> > > >1. create a new directory on any mount
 point, then create
> > > > sub-directories where you will store your
 init, config files, as
> > > > well, as destination directories for
 bdump, udump, cdump and arch.
> > > > you may want to have a create directory
 where you can keep your
> > > > create sql script. Create directory for
 your datafiles.
> > > >
> > > >2. use your appropriate shell command to
 change ORACLE_SID env to
> > > > point to a new SID (let's for an example
 sake call it test1).
> > > >
> > > >3. start up svrmgrl, issue connect internal,
> > > > and now run your sql script to create
 your database
> > > > (i'm not going to go into the semantics
 what tablespaces you need
> > > > etc, i assume you know that).
> > > >
> > > >4. update your listener.ora and tnsnames.ora
 files. you may also want
> > > > to add a link from $ORACLE_HOME/dbs
 directory to point to your
> > > > init file, so next time you want to start
 it you can just startup
> > > > svrmgrl and issue startup command.
> > > >
> > > > (if you need a sql script to create a
 database let us know and we
> > > > will post one for you).
> > > >
> > > >Alex Sokol
> > > >
> > > >In article

 <sfmbnpdo5c246_at_corp.supernews.com>,
> > > > "Bruce" <bey_at_3eg.com> wrote:
> > > >> Hi,
> > > >>
> > > >> I was wondering if somebody can tell me
 how to create a new SID
 for a
 new
> > > >> Database in Linux using RedHat 6.1 and
 Oracle 8.0.5 Server Edition.
 In NT it
> > > >> is easy, but in Linux I am not certain how
 to do it. I was able to
 create a
> > > >> new database beside the default database,
 but the big question is
 how
 to
> > > >> connect to that database without the SID.
> > > >>
> > > >> Thanks in advance.
> > > >>
> > > >> Bruce
> > > >>
> > > >>
> > > >
> > > >
> > > >Sent via Deja.com http://www.deja.com/
> > > >Before you buy.
> > > >
> > >
> > >
> >
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.
> >
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

Sent via Deja.com http://www.deja.com/
Before you buy. Received on Mon Apr 24 2000 - 00:00:00 CDT