Re: oracle 11g dataguard on Netapps
Date: Thu, 12 Feb 2009 12:51:44 +0100
You can install the binaries in NetApp volumes or local disk, it is up to you, but if you have plenty of spaces in your local disks (now days 100 of GB is not unusual) I dont see why you want to install them in a NAS device, I dont see any advantage.
Again I dont have experience with NetApp snapshots but I have used EMC Snapview which should be similar technology, you put your database in backup mode, take the snap and end the backup mode. More or less there are steps to get two controlfile copies and archived logs, probably googling can get you detail steps.
-- LSC On Wed, Feb 11, 2009 at 9:07 PM, Joan Hsieh <joan.hsieh_at_tufts.edu> wrote:Received on Thu Feb 12 2009 - 05:51:44 CST
> Hi Matt,
> Thanks for your prompt reply, what is the disadvantage to install the
> binary on Netapps giving that we have 250gb storage? I am thinking if
> primary and standby shared a same oracle home, it will save time for
> patching or upgrading the binary, besides, it also can take the benefit of
> the snapshots????
> I just know little about NetApps snapshots, how it work for you? do you use
> it for database recovery? Do you need to script the database in hot backup
> mode in order to that? That's why I thought we could benefit from the shared
> oracle home.
> Matthew Zito wrote:
>> I would recommend keeping your binary installs on the local filesystem
>> of your servers - NetApp is expensive storage, and it's useful to keep
>> that install local.
>> As far as separating volumes, you have to remember that on NetApp,
>> having two separate volumes does not necessarily automatically imply
>> that they are on different physical spindles. On NetApp you have
>> something called an "aggregate", which is basically a RAID group (or
>> multiple raid groups, potentially). Out of this aggregate, volumes are
>> carved up, sharing the same physical spindles. The major advantage of
>> having multiple volumes from the same aggregate is administrative - you
>> can make sure that a bad actor on one volume doesn't fill up the whole
>> Also, keep in mind that since NetApp uses a copy-on-write filesystem, it
>> can be useful to share a datafile and its index on the same aggregate,
>> as NetApp will make sure those writes are sequential.
>> For redundancy's sake, if you can spare the disk space, keep at least
>> one set of redo logs, controlfiles, etc. on a separate aggregate. That
>> will help protect you against a multiple disk failure scenario. Also,
>> familiarize yourself with NetApp snapshots, they're very useful.
>> -----Original Message-----
>> From: oracle-l-bounce_at_freelists.org
>> [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Joan Hsieh
>> Sent: Wednesday, February 11, 2009 2:49 PM
>> To: oracle_l
>> Subject: oracle 11g dataguard on Netapps
>> Hi listers,
>> Our new project is creating a 11g dataguard environment on NetApps
>> storage. I have the following questions like to ask;
>> 1. What is the best practice of volume configuration? should we configure
>> single volume for data, single volume for redo?
>> 2. As oracle home, is it better to install on the local file system or on
>> NetApps? If it is best on Netapps, then should the Oracle home be shared
>> with primary and standby sites? Our primary and standby are on separate