Oracle FAQ Your Portal to the Oracle Knowledge Grid

Home -> Community -> Mailing Lists -> Oracle-L -> Re: High-availability database question !!!!

Re: High-availability database question !!!!

From: Arun Annamalai <>
Date: Wed, 25 Jun 2003 10:32:14 -0700
Message-ID: <>

On the database side, The other way to get around this issue is to go for Clustered db environment.
Considering two node cluster, When you apply patch for one database(primary) in cluster, the secondary will be online(as failover from the primary).
This time you can patch the primary. After this is done, Failover the secondary to the primay - but be careful of the data -out of sync.
You need to restore the data to primary from secondary.
The downtime is mandatory and the lenght of the time depends on the size of the database.
(I have seen a BCV with split on EMC disk array -
just datafiles- to be much faster for large databases). Now patch the secondary while primary is online. This time you may have to have both primary and secondary down for restore on the secondary or you can postphone this later. <BLOCKQUOTE
style="PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">

  href="mailto:[EMAIL PROTECTED]">Multiple recipients of list ORACLE-L   

  Sent: Wednesday, June 25, 2003 9:59
  Subject: Re: High-availability database   question !!!!

  1. RAC can duplicate your instances/nodes. But you still have only one database.2. One database means un-availability whenever you have to apply a patch or upgrade the database. No online patches available. No rolling upgrades.3. Dataguard needs to have exactly the same version (down to patch level) on the two (or more) databases. So that just means two databases down if you have to patch or upgrade. That might come as a surprise to some people, but there's basically no 25/8/370 availability with Oracle unless you either specify downtime as something else or don't patch/upgrade at all.4. Logical stuff like replication, logical Dataguard, Shareplex from Quest and others can get around the "everyhing-has-to-be-the-same-version" by using logical replication of some sort - some faster, some slower.MogensJackson Dumas wrote: Goodday all

We have cutomer who is having a database which should be available 24x7.
Now we can't get even a maintenance time to do the database. We tried on several occasion but nothing happens. Now they want us to provide them with a solution to be able to work on "another database" if we need to do maintenance. Could you guys come up with some ideas what we can offer them. Is there any other alternatives besides replication and standby database. Currently the system is running as a package in a cluster environment, HP-UX, so when we go replication route possibilities are that we might have to forget about high-availabilty cluster environment due to things like disk space and memory for switching over during failure. Neither do the standby database a good option in a situation where you'll needing a longer downtime.

Please help.


Junior DBA

On Tue, 24 Jun 2003 08:19:18 -0800 ([EMAIL PROTECTED]) wrote:   

    Hi there,

I got a question on log. If my program crashes, can I check some log

    to see what   

    recent transaction is? It will give me a big help on trouble     shooting.   


Please see the official ORACLE-L FAQ:

Fat City Network Services    -- 858-538-5051
San Diego, California        -- Mailing list and web hosting services
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L

(or the name of mailing list you want to be removed from). You may
also send the HELP command for other information (like subscribing). _______________________________________________________________________ LOOK GOOD, FEEL GOOD - WWW.HEALTHIEST.CO.ZA Cool Connection, Cool Price, Internet Access for R59 monthly @ WebMail
Received on Wed Jun 25 2003 - 12:32:14 CDT

Original text of this message