Re: ASM capacity planning over time

  • From: Kenny Payton <k3nnyp@xxxxxxxxx>
  • To: jeremy.schneider@xxxxxxxxxxxxxx, Joshua Collier <jcoll1970@xxxxxxxxx>
  • Date: Sat, 07 Mar 2015 15:41:33 +0000

I'd recommend adopting a standardized metric collection/graphing/reporting
methodology.  We use Graphite.  Its simple to load metrics from almost
anywhere, including your favorite asm reporting shell script.  Being able
to correlate network traffic, system load, application metrics and database
metrics in a single location is priceless.  There are plenty of modules
available for loading stats, like collectl for Linux, but the beauty of it
for me is you can use whatever you like to format the metrics into name,
date, value and pipe it to nc to push it across your network.

Kenny

On Fri, Mar 6, 2015, 3:47 PM Jeremy Schneider <
jeremy.schneider@xxxxxxxxxxxxxx> wrote:

> cluster asm -> all metrics -> Disk Group Usage...
> - DG Free
> - DG Usable
> - DG Usable Free
> - DG Used %
> - DG Used % Safely Usable
> - Redundancy
> - Size
>
> Joshua - see attached illustration (might get stripped on the mailing list)
>
> --
> http://about.me/jeremy_schneider
>
>
> On Fri, Mar 6, 2015 at 1:10 PM, Joshua Collier <jcoll1970@xxxxxxxxx>
> wrote:
> > so OEM will collect from the asm instance the storage usage and store it
> in
> > a repositoy for display over time?
> >
> > On Wed, Mar 4, 2015 at 11:09 AM, Jeremy Schneider
> > <jeremy.schneider@xxxxxxxxxxxxxx> wrote:
> >>
> >> The "free" version of OEM (haha.. ok it comes with the db licenses)
> >> will do this for you.  I've done that in the past... just have to
> >> build up the courage to tackle OEM.  :)  I don't have experience with
> >> other monitoring systems (quest foglight, solarwinds orion, etc) but
> >> I'd imagine that whichever one you've got, it has this functionality.
> >>
> >> -J
> >> --
> >> http://about.me/jeremy_schneider
> >>
> >>
> >> On Wed, Mar 4, 2015 at 1:21 PM, Joshua Collier <jcoll1970@xxxxxxxxx>
> >> wrote:
> >> > oracle version 12c
> >> > ASM instance has only x$ and V$ tables. It has no AWR.
> >> >
> >> > How do people compute growth rate over time for all storage managed
> by a
> >> > particular ASM instance?
> >> >
> >> > In an environment where PDB are self service and dropped and added
> >> > regularly, historical growth reporting at the database level won't
> >> > accurately capture usage.
> >> >
> >> > Thanks for your time!
> >> >
> >> > Josh C>
> >
> >
>

Other related posts: