Re: block chg tracking

From: kyle Hailey <>
Date: Tue, 3 Feb 2015 22:06:09 -0800
Message-ID: <>

here's a short blog post on BCT setup and overhead

we use BCT extensively at Delphix to efficiently collect changed blocks from our last RMAN incremental SCN in order to build fast provisioning points for thin clones a.k.a virtual databases. Using a block sharing + redirect on write file systems along with compression and the changed blocks we can provide access to a database any where in the past 30 days in a few minutes for typically just over the size of the original database.


On Tue, Feb 3, 2015 at 7:18 AM, Jeremy Schneider <> wrote:

> I've used BCT in a number of different environments, lots of 11g, and
> haven't seen problems. I have heard about the bugs in some older
> versions but personally haven't hit them (or they were already patched
> on environments I worked with). I haven't noticed significant
> overhead; in my opinion, any overhead that does exist along with the
> space used is very much worth the time and cpu and i/o savings during
> incremental backups.
> -J
> --
> On Mon, Feb 2, 2015 at 3:11 PM, Cee Pee <> wrote:
> > List,
> >
> > I am considering turning on bct for one of our prod databases. Per manual
> > there will be an increase in the resource usage during regular
> operations if
> > bct is turned on. is there a way to quantify the increase. 10g/11g EE on
> > AIX; not too many simultaneous users. Can 'block changes' and/or 'redo
> size'
> > in awr over the standard period of operation help.
> >
> > CP.
> --

Received on Wed Feb 04 2015 - 07:06:09 CET

Original text of this message