Re: Rebooted servers and then ASM does not bring up

From: fairlie rego <fairlie_r_at_yahoo.com>
Date: Thu, 27 Mar 2008 08:18:35 +1100 (EST)
Message-ID: <407030.33414.qm@web31910.mail.mud.yahoo.com>

Applying the patch is not going to fix the corruption. It is only going to prevent it in the future. This is not the only ASM corruption issue in 10.2.0.3    

  The copy on the NAS existed before this problem.

Jeffery Thomas <jeffthomas24_at_gmail.com> wrote:   Did you apply the patch for this bug? Wouldn't that be sufficient to avoid having to
go the NAS route?

Thanks,
Jeff

On Wed, Mar 26, 2008 at 3:58 AM, fairlie rego wrote:
> Hi,
>
> Coincidentally it was the FRA (FLASH_DG) in my case too. Also the customer
> who initially hit the problem for which the bug was logged also had the
> corruption in their FRA. So I did what you did and just lost some
> archivelogs, some flashback logs and the BCT file. You have to re-create the
> diskgroup (unless someone from Support can use kfed to repair the
> corruption)
>
> But I have had other corruptions in the DATA_DG (the diskgroup holding the
> entire database) in which a restore and recover is the only option assuming
> you cannot switch to a copy in the FRA.
>
> To avoid these single point of failure we hold a second copy of the database
> on a NAS mountpoint.
>
>
> Thanks
> Fairlie
>
> Edgar Chupit wrote:
> Dear Fairlie,
>
> Thank you for sharing this information, it's very useful for analyzing
> current risks.
>
> Can you, please, also share what type of diskgroup have you lost, was
> it flash recovery area disk group as in our case, or it was a
> diskgroup with the data? How did you resolved this situation, did you
> recreated diskgroup and restored/recovered lost datafiles or have you
> used another approach.
>
> Thanks.
>
> On Tue, Mar 25, 2008 at 11:01 PM, fairlie rego wrote:
> > Coincidentally I hit the same issue a couple of weeks ago.
> >
> > The corruption is noticed during a reconfiguration triggered when an ASM
> > instance joins or leaves the cluster. It is seen in the Active Change
> > Directory (file#3) which is the part of the ASM disk group analogous to
> the
> > RDBMS redo stream which is used for instance recovery.
> >
> > Fairlie Rego
> > Senior Oracle Consultant
> > http://el-caro.blogspot.com/
> > M: +61 402 792 405
> >
>
>
> --
> Best regards,
> Edgar Chupit
> callto://edgar.chupit
>
>
>
>
>
>
>
>
> Fairlie Rego
> Senior Oracle Consultant
> http://el-caro.blogspot.com/
> M: +61 402 792 405
>
>
>
>
> ________________________________
> Get the name you always wanted with the new y7mail email address..

          Fairlie Rego
Senior Oracle Consultant
  http://el-caro.blogspot.com/
  M: +61 402 792 405            

---

Get the name you always wanted with the new y7mail email address..

--
http://www.freelists.org/webpage/oracle-l

Received on Wed Mar 26 2008 - 16:18:35 CDT