Oracle FAQ Your Portal to the Oracle Knowledge Grid

Home -> Community -> Usenet -> c.d.o.server -> Re: Oracle Database 10g R2 and RHEL4 Update 3

Re: Oracle Database 10g R2 and RHEL4 Update 3

From: Andreas Sheriff <>
Date: Sun, 7 May 2006 13:45:22 -0700

<div class="mail">
<address class="headers">
<span id="from">
<dfn>From</dfn>: Andreas Sheriff <>
<span id="date"><dfn>Date</dfn>: Sun, 7 May 2006 13:45:22 -0700</span>
<span id="message-id"><dfn>Message-ID</dfn>: <X%s7g.1054$KB.819_at_fed1read08>


"maxim2k" <maxim2k_at_gmail&#46;<!--nospam-->com> wrote in message news:e8baa$445dd246$5277c1be$13768_at_news0&#46;<!--nospam-->
<i class="quotelev1">> On 5/3/06 8:56 AM, Andreas Sheriff wrote:</i><br>
<i class="quotelev2">> > Check out ASM. It's very, very cool.</i><br>
<i class="quotelev1">></i><br>
<i class="quotelev1">> Can you quickly summarize the advantages over regular storage?</i><br>
<i class="quotelev1">></i><br>
<i class="quotelev1">> Thanks.</i><br>

First, and most important, IMHO, it takes the control of your database files out of the hands of the sysadmin and people who are knowledgeable enough to be dangerous. I once had a fellow programmer delete an 8i datafile because he thought it wasn't needed anymore; aargh!

Next, ASM automatically rebalances and effectively stripes over all disks and makes redundant the data through failgroups. If a disk fails, the data contained on that disk is redistributed, rebalanced, and made redundant over surviving disk. EM should email you of the failure. All you would have to do is to remove the failed disk from the diskgroup and optionally from ASMLIB, remove the failed disk from the OS, physically remove the failed disk, replace the failed disk with a good disk, have the OS recognize the good disk, optionally add the disk to ASMLIB, and finally add the disk to the diskgroup / failgroup. ASM will then automatically rebalance and make redundant data over this new disk. Say, I wonder if one day vendors will implement ASM on their HBA? That would be nice...

Third, ASM is a perfect datastore for OMF.

Fourth, adding storage to ASM is easy and managed entirely by Oracle. No more vendor FS foulups. When you add new disks to a diskgroup, ASM automatically rebalances data over these new disks so throughput can be increased just by addding new disks and a new pipe.

Fifth, if you're doing RAC with SE, you can *only* use ASM.

There are many more benefits to ASM. If you want a more exhausive list, I suggest that you read up on ASM on Oracle's web site and search the web for references on ASM.

Note, though, that there have been studies comparing ASM to RAID (software and hardware). Studies show that for single instance, RAID is a bit faster than ASM for file access, but for RAC, ASM shines.

One thing that I don't like about ASM, though, is that there is no *direct* access to datafiles. I suppose this was a planned *feature*. But with all the benefits of ASM I'm willing to overlook that.

Perhaps someday someone will come up with an OS accessable ASMFS, but then we'll run into the same problem with fellow programmers and sysadmins moving, removing, and corrupting datafiles.

If you need someone to help you set up ASM I'd be glad to consult.

Contact me using my regular email.

The format is:



<myfirstinitial> is a variable to be replaced with "a", without the quotes.
<mylastname> is a variable to be replaced with "sheriff", without the


Andreas Sheriff
Oracle 9i Certified Professional
Oracle 10g Certified Professional
Oracle 9i Certified PL/SQL Developer
"If you don't eat your meat, you cannot have any pudding.
"How can you have any pudding, if you don't eat your meat?"

Reply only to the group.

<span id="received"><dfn>Received on</dfn> Sun May 07 2006 - 15:45:22 CDT</span>
Received on Sun May 07 2006 - 15:45:22 CDT

Original text of this message