RE: Direct NFS and ZS3

From: Hameed, Amir <Amir.Hameed_at_xerox.com>
Date: Wed, 6 May 2015 19:15:17 +0000
Message-ID: <AF02C941134B1A4AB5F61A726D08DCED0E1B9F02_at_USA7109MB012.na.xerox.net>



Kevin,
Are there any new dNFS features or enhancements introduced in 12c? I am particularly interested for storage agnostic features as we use EMC's NAS.

Amir
-----Original Message-----
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Kevin Jernigan Sent: Wednesday, May 06, 2015 3:02 PM
To: dmarc-noreply_at_freelists.org; oracle-l_at_freelists.org Subject: Re: Direct NFS and ZS3

Mldaen,

The primary design goal for Direct NFS (dNFS) was / is to provide SAN-equivalent (or better) performance, in terms of both latency and throughput, while using NFS / Ethernet infrastructure. A secondary goal is to simplify the configuration and tuning process for Oracle Database with NFS storage. dNFS accomplishes these goals by implementing the NFS client inside Oracle Database, rather then using the OS-supplied "kernel" NFS client. This allows dNFS to skip some parts of the networking stack, and to skip some of functionality that is required for a general-purpose NFS client, such as write ordering. In addition, dNFS creates a separate connection to the NFS server for each Oracle process, unlike kNFS, which essentially multiplexes all the processes' I/O's through one or a very small number of connections to the NFS server. There are other optimizations in dNFS which provide major performance improvements over kNFS, and which allow dNFS to auto-configure itself based on interrogating the NFS server.

In general, if the NFS server can handle the workload, then dNFS can provide SAN or iSCSI-equivalent performance, with very little configuration work required of the DBA or system administrator.

More specifically, with Oracle ZFS storage such as ZS3 and ZS4, dNFS includes additional info with each I/O, which allows the ZS storage to auto-configure to provide ideal performance for Oracle Database I/O. This feature is called Oracle Intelligent Storage Protocol - OISP - and is available starting with Oracle Database 12c.

Bottom line: I know of no good reason to use kNFS instead of dNFS when setting up Oracle Database with NFS storage, and there are lots of good reasons to consider dNFS versus iSCSI or FC storage.

-KJ

--
Kevin Jernigan
Senior Director Product Management
Advanced Compression, Hybrid Columnar
Compression (HCC), Database File System (DBFS), SecureFiles, Database Smart Flash Cache, Total Recall, Database Resource Manager (DBRM), Direct NFS Client (dNFS), Continuous Query Notification (CQN), Index Organized Tables (IOT), Information Lifecycle Management (ILM) +1-650-607-0392 (o)
+1-415-710-8828 (m)

On 5/6/15 12:49 AM, Mladen Gogala wrote:
> On 05/05/2015 12:16 PM, Fernando Jose Andrade wrote:
>> I don’t know if this configuration is worst than DNFS, I have read
>> around the web that DNFS beats iSCSI
> That depends on whether you will utilize iSCSI HBA's like QLogic 4K or
> not. With HBA, I find the iSCSI performance better than NFS. I have no
> numbers to share, only an impression. Also, here is a nice little
> comparative write-up:
>
> http://www.infoworld.com/article/2616802/infrastructure-storage/your-f
> ateful-decision--nfs-or-iscsi-.html
>
>
> As far as snapshots are considered, ZFS does CoW snapshot, which will
> very quickly kill the performance of your appliance. So, since ZS3
> already offers file system and is not LUN based, like EMC or NetApp, I
> would go with NFS which doesn't require additional pieces of HW (iSCSI
> HBA) but would refrain from doing snapshots. As for the choice
> between the kernel NFS and dNFS, I usually go with kernel, which is
> more standard to configure.
>

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

i0zX+n{+i^ Received on Wed May 06 2015 - 21:15:17 CEST

Original text of this message