NetApp - Databases

Thank you Oracle Magazine Team

Hi All,

This is neto from Brazil (in RTP) :-)

How are you?

I would like to thank you very much Oracle Magazine Team for the publication about myself and NetApp on Efficiency Experts session on Oracle Magazine - September/October 2010.

Oracle Magazine September/October 2010:

http://www.oraclemagazine-digital.com/oraclemagazine/20100910?sub_id=CD0hir2A2Hvb&u1=01#pg2

Efficiency Experts session

http://www.oraclemagazine-digital.com/oraclemagazine/20100910?sub_id=CD0hir2A2Hvb&u1=01#pg39

Thank you very much for all your support and kindness. I really appreciate.

All the best

neto

Net(App)O(racle)

Using Crash-Consistent Snapshot Copies as Valid Oracle Backups

Hi All,

This is neto from Brazil

How are you?

http://media.netapp.com/documents/tr-3858.pdf

This document addresses the use of NetApp® Snapshot™ technology as a method of capturing Oracle® backups in light of Oracle’s recent announcement of official support for using third-party snapshot technologies to create crash-consistent images without backup mode. Using the consistency group feature of Data ONTAP®, customers can capture cross-volume Snapshot copies that preserve write ordering, which is crash consistent and can be used to create recoverable backup images. This document also contains NetApp’s best practices for Oracle Databases to prevent blocks from fracturing in the I/O infrastructure.

I wish you a good reading for this weekend :-)

All the best

neto

NetApp – I love this company!

101 ASM

Hi All,

This is neto from Brazil

How are you?

Some useful ASM commands:

 

[oracle@atl46003][+ASM][~]$ rpm -qa | grep oracleasm

oracleasm-support-2.1.2-1.el5

oracleasm-support-2.1.3-1.el5

oracleasm-2.6.18-92.el5-2.0.5-1.el5

oracleasmlib-2.0.4-1.el5

oracleasm-2.6.18-53.el5-2.0.4-1.el5

oracleasmlib-2.0.3-1.el5

 

[root@atl46003 ~]# lsmod | grep asm

oracleasm              84136  1

 

[root@atl46003 ~]# service oracleasm status

Checking if ASM is loaded: yes

Checking if /dev/oracleasm is mounted: yes

 

[root@atl46003 ~]# service oracleasm listdisks

ARCH1

DATA1

DATA2

LOG1

 

[root@atl46003 ~]# service oracleasm querydisk DATA1

Disk "DATA1" is a valid ASM disk on device [253, 12]

 

[root@atl46003 ~]# ls -la /dev/mapper | grep 253 | grep 12

brw-rw----  1 root disk 253, 12 Mar 13 00:58 360a98000486e2f4c5334555a30504934p1

 

[root@atl46003 ~]# hexdump -C /dev/mapper/360a98000486e2f4c5334555a30504934p1 | head -3

00000000  01 82 01 01 00 00 00 00  00 00 00 80 6a 17 59 3e  |............j.Y>|

00000010  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|

00000020  4f 52 43 4c 44 49 53 4b  44 41 54 41 31 00 00 00  |ORCLDISKDATA1...|

 

[root@atl46003 ~]# ls /dev/oracleasm/disks/DATA1 -la

brw-rw---- 1 oracle dba 253, 12 Mar 13 01:00 /dev/oracleasm/disks/DATA1

 

[root@atl46003 ~]# service oracleasm querydisk ARCH1

Disk "ARCH1" is a valid ASM disk on device [253, 13]

 

[root@atl46003 ~]# ls -la /dev/mapper | grep 253 | grep 13

brw-rw----  1 root disk 253, 13 Mar 13 00:58 360a98000486e2f4c5334555a306c3332p1

 

[oracle@atl46003][prod][~]$ export ORACLE_SID=+ASM

 

[oracle@atl46003][+ASM][~]$ sqlplus / as sysdba

SQL*Plus: Release 10.2.0.4.0 - Production on Wed Apr 7 17:42:45 2010

Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.

Connected to:

Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bit Production

With the Partitioning, OLAP, Data Mining and Real Application Testing options

SYS@+ASM SQL> set lines 100

SYS@+ASM SQL> select group_number, name, state, total_mb, free_mb, type from v$asm_diskgroup;

GROUP_NUMBER NAME                           STATE         TOTAL_MB    FREE_MB TYPE

------------ ------------------------------ ----------- ---------- ---------- ------

           1 ARCH                           MOUNTED          51197      46792 EXTERN

           2 DATA                           MOUNTED         102394     101050 EXTERN

           3 LOG                            MOUNTED          10233       9990 EXTERN

SYS@+ASM SQL>  select group_number,disk_number,name,state,mode_status,total_mb,free_mb from v$asm_disk;

GROUP_NUMBER DISK_NUMBER NAME                           STATE    MODE_ST   TOTAL_MB    FREE_MB

------------ ----------- ------------------------------ -------- ------- ---------- ----------

           1           0 ARCH1                          NORMAL   ONLINE       51197      46792

           2           0 DATA1                          NORMAL   ONLINE       51197      50521

           2           1 DATA2                          NORMAL   ONLINE       51197      50529

           3           0 LOG1                           NORMAL   ONLINE       10233       999

SYS@+ASM SQL> set lines 100

SYS@+ASM SQL>  col path format a20

SYS@+ASM SQL> select group_number,disk_number,name,state,mode_status,path from v$asm_disk;

GROUP_NUMBER DISK_NUMBER NAME                           STATE    MODE_ST PATH

------------ ----------- ------------------------------ -------- ------- --------------------

           1           0 ARCH1                          NORMAL   ONLINE  ORCL:ARCH1

           2           0 DATA1                          NORMAL   ONLINE  ORCL:DATA1

           2           1 DATA2                          NORMAL   ONLINE  ORCL:DATA2

           3           0 LOG1                           NORMAL   ONLINE  ORCL:LOG1

SYS@+ASM SQL> alter diskgroup DATA add disk ‘ORCL:DATA3’;

Diskgroup altered.

SYS@+ASM SQL> alter diskgroup DATA drop disk ‘DATA1’ rebalance power 0;

Diskgroup altered.

SYS@+ASM SQL> alter diskgroup DATA check all repair;

Diskgroup altered.

SYS@+ASM SQL> alter diskgroup DATA rebalance power 11;

Diskgroup altered.

SYS@+ASM SQL> select * from v$asm_operation;

GROUP_NUMBER OPERA STAT      POWER     ACTUAL      SOFAR   EST_WORK   EST_RATE EST_MINUTES

------------ ----- ---- ---------- ---------- ---------- ---------- ---------- -----------

           2 REBAL REAP         11          9         56       1047       2729           0

 

All the best

neto

NetApp – I love this company!

Using Iometer to Generate I/O

Hi All,

This is neto from Brazil

How are you?

Iometer is both a workload generator (that is, it performs I/O operations in order to stress the system) and a measurement tool (that is, it examines and records the performance of its I/O operations and their impact on the system). It can be configured to emulate the disk or network I/O load of any program or benchmark, or can be used to generate entirely synthetic I/O loads. It can generate and measure loads on single or multiple (networked) systems.

Iometer can be used for measurement and characterization of:

• Performance of disk and network controllers.
• Bandwidth and latency capabilities of buses.
• Network throughput to attached drives.
• Shared bus performance.
• System-level hard drive performance.
• System-level network performance.

Iometer consists of two programs, Iometer and Dynamo.

• Iometer is the controlling program. Using Iometer’s graphical user interface, you configure the workload, set operating parameters, and start and stop tests. Iometer tells Dynamo what to do, collects the resulting data, and summarizes the results in output files. Only one copy of Iometer should be running at a time; it is typically run on the server machine.
• Dynamo is the workload generator. It has no user interface. At Iometer’s command, Dynamo performs I/O operations and records performance information, then returns the data to Iometer. There can be more than one copy of Dynamo running at a time; typically one copy runs on the server machine and one additional copy runs on each client machine.

Dynamo is multithreaded; each copy can simulate the workload of multiple client programs. Each running copy of Dynamo is called a manager; each thread within a copy of Dynamo is called a worker.

Here you have a video explaining How to Generate I/O using Iometer

In the video, basically we have:

• GUI – Data Collector – running on Windows
• Dynamo – workload generator – running on Linux 64 bits
• 3 Servers running Linux RedHat 5 Update 4 (64 bits) generating workload (dynamo) consolidating all results to GUI (Windows)

Each Server has 2 workers (threads) with 128 outstanding I/Os each one accessing the following disks:

• Worker 1 – 128 outstanding I/Os against dm-25 (multipathing device located at NetApp Controller 1 with 4 shelves – 300GB FC disks 15K RPM) using FC 4Gbps
• Worker 2 - 128 outstanding I/Os against dm-26 (multipathing device located at NetApp Controller 2 with 3 shelves – 300GB FC disks 15K RPM) using FC 4Gbps

Also I have used the disks directly (no partitioning) to avoid misaligned partitions and filesystem caching.

The workload was aimed at simulating random access with 100% write operations:

• Transfer Request Size – 4096 bytes
• Percent Random/Sequential Distribution – 100% Random
• Percent Read/Write Distribution – 100% Write
• Align I/Os on – 4096 bytes

The parameter “Align I/Os on” is commonly mistakenly used in the field. By default, Iometer has a default parameter for “Align I/Os on” – Sector Boundaries (512 bytes).

This morning, I was talking with my big friends John Elliott, Eric Forgette and Peter Learmonth about it. When using enterprise arrays the disk, called a LUN, is often a logical representation of many physical disks  (in my case I have a LUN spread out in more than 50 disks).  Since enterprise arrays are organized using larger groups of bytes than sectors (512 bytes), testing with sector sized IO will result in faulty results.  In addition, most modern filesystems and databases work in groups of bytes greater than or equal to 4096 bytes.

The smallest atomic unit a storage device can access is called a sector. With very few exceptions, a sector size of 512 bytes has been akin to a mathematical constant in the storage industry for decades. That picture is now rapidly changing with hard drives moving to 4KB sectors. Flash-based solid state drives and enterprise RAID arrays also have alignment and block size requirements above and beyond what we have traditionally been honoring.

This is an excerpt of a great paper about I/O topology: I/O Topology by Martin K. Petersen (Oracle).

There is also an interesting sentence on the paper that goes:

If the operating system submits a request smaller than 4KB, or if the request submitted is misaligned and straddles two physical blocks, the drive firmware will have to perform a read-modify-write cycle. This incurs a significant performance penalty as the drive will have to perform an extra platter rotation.

That is the reason why it makes sense to configure the workload generator (Iometer) to use 4KB or multiple of 4KB on the “Aligned I/O on” field to avoid generating extra I/O for nothing.

So, after following such best practices, you just need to apply the “Access Specification” profile to all workers and click Start to begin the test.

If you have any questions please feel free to contact me.

Hope this helps

All the best

neto

NetApp – I love this company!

Pele or Maradona? Brazil vs Argentina? Isn’t it Better “One Team”?

Hi All

This is neto from Brazil

How are you?

In the football world, there are big discussions about how is the best soccer player of the world?

Diego Armando Maradona "Hand of God" against England (World Cup of 1986), while the second goal was a spectacular 60-meter weave through six England players, commonly referred to as "The Goal of the Century".

Maradona at FIFA - FIFA Goal of the Century winner (for his second goal against England in 1986 FIFA World Cup quarter-final)

or

Edison Arantes do Nascimento (Pelé) simply in his career he played in 1,363 matches and scored 1,282 goals. Simply Pelé

Pelé at FIFA - FIFA Player of the Century and FIFA Order of Merit: awarded in 2004
MasterCard Team of the Century inductee International Olympic Committee's Athlete of the Century

So… Argentina vs Brazil?

Isn’t it better “One Team”?

Last week I had the pleasure to participate of a big and complex POC in Mendoza Argentina doing tests with Oracle and NetApp.

I’m very thankful to be part of the team with fantastic players.

Here is the picture

Matias (Oracle Partner Consulting), neto from Brazil (NetApp), Mariano Pipino (customer) and David Cordero (NetApp Chile)

In 4 days (more than 16 hours per day), only a fantastic team could make that happen. More than 100.000 IO per second, Oracle RAC 3 nodes, ASM, Oracle tuning and huge improvements in all results.

I would like to thank you very much for all kindness and all support  and I hope to have contributed with the outstanding results (like Maradona and Pelé) in Mendoza.

BTW, my dear “Hermanos” please I could not resist myself, for me no doubts…

Pelé is always the BEST :-)

Un abrazo,

All the best

neto

NetApp – I love this company!

Technical Report: Providing Zero Downtime for Enterprise Applications Using NetApp and Oracle

Hi All,

This is neto from Brazil

How are you?

It is my pleasure to announce the Technical Report: Providing Zero Downtime for Enterprise Applications Using Oracle Real Application Cluster on Extended Distance Cluster, Oracle Automatic Storage Management Normal Redundancy, and NetApp MetroCluster.

http://media.netapp.com/documents/tr-3816.pdf

Whether you have a single data center, a campus, or a metropolitan-wide environment, a combined solution using NetApp and Oracle technologies is a cost-effective alternative that can provide continuous data availability for your mission-critical enterprise applications. NetApp MetroCluster is an industry-leading solution that combines NetApp storage array–based clustering and synchronous mirroring to help deliver continuous availability and minimal data loss. NetApp Fabric MetroCluster combined with Oracle Real Applications Clusters (RAC) on Extended Distance Clusters, and Oracle Automated Storage Management (ASM) Normal Redundancy offers transparent recovery from failures so mission-critical Oracle databases can continue uninterrupted with zero downtime.

This paper combines the robustness of NetApp and Oracle technologies.

Enjoy the reading!

All the best

neto

NetApp – I love this company!

Video Demo of upcoming TR about ZERO downtime for Enterprise Applications using NetApp and Oracle technologies

Hi All

This is neto from Brazil

How are you?

Does your business need a solution to provide ZERO downtime for Enterprise Applications using Oracle?

The solution is very simple: Use NetApp!

Learn how watching the video below:

Note: if you would like to see the video in Full Screen format please click here

The TR is coming soon… Stay tuned!

All the best

neto

NetApp – MY LIFE!

ASM Intelligent Data Placement – We have WAFL

Hi All

This is neto from Brazil

How are you?

I have the pleasure to introduce my friend and brother Neil Gerren. Neil is one the best Oracle Gurus that I have worked in all my life.

Yesterday, we were talking about ASM Intelligent Data Placement (new feature in Oracle 11g R2 – ASM) and we would like to talk a little more about this …

This feature is presenting itself more frequently with our customers.  First, understand that it is Oracle's assumption ASM is running on DAS (Direct Attached Storage,) and JBOD (Just a Bunch of Disks.)  The goal of ASM is to improve manageability, data protection, and throughput while reducing latency as well as possibly eliminating the need for an "expensive" disk array.  It adds intelligence to "dumb" storage.  However, it is no substitute for modern enterprise storage systems.

If you are not taking advantage of a modern storage array, ASM Intelligent Data Placement can help you leverage the fact that data written on the outside of the platter will retrieve faster, and with lower latency than data residing on the inside.  It is the first "intelligence" feature added to ASM, and I suspect we can look forward to more!

Now extend this thinking to Oracle.  What if the database had a way of segregating your writes so that tablespaces requiring higher performance were actually written to the outside of the disk while tablespaces with lower required service times are written to the inside of the disk.

So with 11gR2 ASM, you can categorize your data files as "HOT" and "COLD".  This is done with a simple command:

    Alter diskgroup D0001 modify file '+D0001/DBPRD01/DATA/SYSADM01.256.123456789' attributes HOT;

    -or-

    Alter diskgroup D0001 modify file '+D0001/DBPRD01/DATA/GL_IDX01.256.123456789' attributes COLD;

After changing these attributes, ASM will make Zoned Bit Recording calls to place the data on disk according to this new attribute.  Keep in mind that there will be significant head movement between the hot and cold zones on disk.

By virtue of the fact that you are reading a Blog written by yours truly, an employee of an enterprise storage company, you probably prefer to use a modern storage array in lieu of JBOD.  Disk arrays provide much better reliability availability than JBOD.  They also allow you to better share storage resources across your applications and your data centers.  Consider also today's storage system's ability to monitor and manage the health of disks, enabling you to be proactive in keeping your data online and accessible.  Disk failures are handled transparently and automatically.  However, the greatest and most obvious advantage is that storage maintenance activities (i.e.: disk rebalancing) present no load whatsoever to your database servers.

In the case of NetApp's ONTAP operating system, and particularly, our ability to aggregate your database workload across many spindles, we have a lot more R&D invested towards intelligently writing data to disk than Oracle's ASM.  Some may be confused by the term WAFL (Write Anywhere File Layout) and for good reason!  We don't actually write "anywhere."  We write to the best location in order to give you the best performance, minimizing head movement while grouping like data together.  It is our R&D investment that results in storage that is much easier to manage.  Consider lastly, NetApp's ability to create thin database clones in minutes, and a pure ASM/JBOD solution becomes mundane and potentially very costly to maintain.

All the best

neto

NetApp – I love this company!

How to convert an SCN to a date in Oracle 10g

Hi All,

This is neto from Brazil

How are you?

Here is an useful info to convert an SCN to a date in Oracle 10g. You can use this conversion for all restore’s process.

All the best

neto

NetApp – I love this company!

Renaming ASM Diskgroup on Oracle 10g

Hi All,

This is neto from Brazil

How are you?

Bye Bye kfed

Rename diskgroup (renamedg) is a new feature of 11gR2 .
We can use this tool to rename your 10g or 11gR1 ASM diskgroup. Basically, we need to just install 11gR2 SIHA Grid Infrastructure software only installation.

Dismounting the diskgroup

# asmcmd umount DATA

Checking if diskgroup was dismounted:

# asmcmd lsdg
State Type Rebal Sector Block AU Total_MB Free_MB Req_mir_free_MB Usable_file_MB Offline_disks Voting_files Name
MOUNTED NORMAL N 512 4096 1048576 32756 31828 244 15792 0 N DATA_OCR/
MOUNTED EXTERN N 512 4096 1048576 16378 10134 0 10134 0 N LOBDATA/


Using renamedg

# renamedg phase=both dgname=DATA newdgname=DG_PROD_DATA verbose=true
Parsing parameters..
Parameters in effect:
Old DG name : DATA
New DG name : DG_PROD_DATA
Phases :
Phase 1
Phase 2
Discovery str : (null)
Clean : TRUE
Raw only : TRUE
renamedg operation: phase=both dgname=DATA newdgname=DG_PROD_DATA verbose=true
Executing phase 1
Discovering the group
Performing discovery with string:
Identified disk ASM:/opt/oracle/extapi/64/asm/orcl/1/libasm.so:ORCL:DATA1 with disk number:0 and timestamp (32925601 -972709888)
Identified disk ASM:/opt/oracle/extapi/64/asm/orcl/1/libasm.so:ORCL:DATA2 with disk number:1 and timestamp (32925601 -972709888)
Identified disk ASM:/opt/oracle/extapi/64/asm/orcl/1/libasm.so:ORCL:DATA3 with disk number:2 and timestamp (32925601 -972709888)
Checking for hearbeat...
Re-discovering the group
Performing discovery with string:
Identified disk ASM:/opt/oracle/extapi/64/asm/orcl/1/libasm.so:ORCL:DATA1 with disk number:0 and timestamp (32925601 -972709888)
Identified disk ASM:/opt/oracle/extapi/64/asm/orcl/1/libasm.so:ORCL:DATA2 with disk number:1 and timestamp (32925601 -972709888)
Identified disk ASM:/opt/oracle/extapi/64/asm/orcl/1/libasm.so:ORCL:DATA3 with disk number:2 and timestamp (32925601 -972709888)
Checking if the diskgroup is mounted
Checking disk number:0
Checking disk number:1
Checking disk number:2
Checking if diskgroup is used by CSS
Generating configuration file..
Completed phase 1
Executing phase 2
Looking for ORCL:DATA1
Modifying the header
Looking for ORCL:DATA2
Modifying the header
Looking for ORCL:DATA3
Modifying the header
Completed phase 2
Terminating kgfd context 0x2af21517e0a0


Mounting the diskgroup

# asmcmd mount DG_PROD_DATA


# asmcmd lsdg
State Type Rebal Sector Block AU Total_MB Free_MB Req_mir_free_MB Usable_file_MB Offline_disks Voting_files Name
MOUNTED NORMAL N 512 4096 1048576 32756 31828 244 15792 0 N DATA_OCR/
MOUNTED EXTERN N 512 4096 1048576 24567 14267 0 14267 0 N DG_PROD_DATA/
MOUNTED EXTERN N 512 4096 1048576 16378 10134 0 10134 0 N LOBDATA/

I will work on a procedure using FlexClone and I will make a new post with the procedure.

All the best

neto

NetApp – My REASON to LIVE!