Re: Oracle Active Data Guard

From: Noons <wizofoz2k_at_gmail.com>
Date: Wed, 7 Dec 2011 17:43:30 -0800 (PST)
Message-ID: <cc5dcded-c2ac-4b86-99e3-06ce150dbb05_at_o11g2000prg.googlegroups.com>



On Dec 2, 5:36 am, Mladen Gogala <gogala.mla..._at_gmail.com> wrote:
> Oracle 11g has a very interesting option: active DG. Basically, one can
> create a physical standby db which is open for querying. It is much
> better thing than a logical standby database because there is no hassle
> with streams and unsupported types. The cost of the "active DG" is the
> same as the cost of an EE license on the CPU basis, which means that this
> is the same thing as logical standby, at least money wise. There is even
> a new status field:
>
> DGMGRL> show database stby
>
> Database - stby
>
>   Role:            PHYSICAL STANDBY
>   Intended State:  APPLY-ON
>   Transport Lag:   0 seconds
>   Apply Lag:       0 seconds
>   Real Time Query: ON
>   Instance(s):
>     O11
>
> Database Status:
> SUCCESS
>
> The "Real Time Query: ON" means that this database is open for querying.
>
> This is a great thing, especially when paired with fast fail-over
> mechanism. This makes a logical standby completely obsolete. I am getting
> more and more enthusiastic about Oracle 11g. I believe it's the time to
> convert.
>
> --http://mgogala.byethost5.com

And with the DGBroker, it's child's play to switch roles across sites. Flicking the DR site to active and back to inactive is a joy to watch. Bugs not-withstanding, it's finally a good and usable technology. The main reason I'm going 11gr2 next year.

Now, if only they fixed their transmit mechanism to use SAN dark fibre for replication (or any other transmit mechanism - including carrier pigeons!) instead of TCP/IP only.... It'd really be something I could use tomorrow.

Right now, it takes 30 secs to transmit one of my 1GB redo logs to DR via TCP/IP, while the db generates one of them every 20 secs at peak times: hardly a feasible solution, isn't it? Received on Wed Dec 07 2011 - 19:43:30 CST

Original text of this message