Transcript Oracle-ASM-vs-HP-UX-LVM-DusanBaljevic

HP-UX LVM,
OnlineJFS and
Oracle ASM Basics*
Dusan Baljevic
Sydney, Australia
2009
Disk Partitioning Concepts – HP-UX
Partitions can be configured using:
•
The Whole Disk Approach (no volume manager).
•
Logical Volume Manager (LVM).
•
Veritas Volume Manager (VxVM).
Whole Disk Partitioning Concepts
The whole disk approach supports partitioning
a disk in five different ways:
File System
Raw Space
Swap Space
File System
Boot Area
File System
Swap
Swap
Whole Disk Partitioning - Pros
• Simple to use, almost no Unix knowledge
required.
• No licensing.
• Supports any type of physical volume.
Whole Disk Partitioning - Cons
• Partitions cannot span multiple disks.
• Each disk can contain at most one file system
partition.
• Partitions cannot be easily extended.
Logical Volume Manager Concepts
• HP-UX LVM is much more flexible than the whole disk approach.
Introduced in HP-UX 9.0:
– Partitions/volumes can span multiple disks
– Multiple partitions/volumes may be configured on a single disk
– Partitions/volumes can be easily extended and reduced as needs
change
• LVM is included in all current versions of HP-UX
– BaseLVM is included with the operating system
– LVM MirrorDisk/UX is available for an extra charge or in higher OE
releases
Physical
Volume
Physical
Volume
Volume
Group
Logical Volumes
Logical Volume
Manager (LVM)
MirrorDiskUX
(MDUX)
Veritas Volume
Manager (VxVM)
Cluster Volume
Manager (CVM)
Included
in VSE,
HA, and
DCOE
File System Solutions
Volume Management Solutions
HP-UX File Systems and Volume
Managers
Base File System
(VxFS-lite)
MirrorDiskUX
(MDUX)
Cluster File
System (CFS)
Include
d in
Base
OE
Include
d in
VSE,
HA, and
DCOE
HP-UX with Symantec Releases
HP-UX OS Version
Symantec Release
Version
HP-UX 11.23
3.5
4.1
HP-UX 11.31
5.0
4.1
5.0
5.0.1 Sep
2009
Base Products
Base-VxFS
Base-VxFSBase-VxFSIn BaseOS Base-VxFS
Base-VxFS
50
5.0
Base-VxVM
Base-VxVM Base-VxVM
Base-VxFS5.0
Base-VxVMBase-VxVM- Base-VxVMBase-VxVM
50
5.0
5.0
Standalone Products
Online-JFS
B3929DA
B3929EA
B3929FA
B3929EA
B3929FB
B3929FB
VxVM-FULL
B9116AA
B9116BA
B9116CA
B9116BA
B9116CB
B9116CB
Update 3
enhancement
Scalability: LVM Supported maximums
HP-UX 11i v1/v2
LVM
Current L1 Limits
HP-UX 11i v3 (L2 Layout Limits)
On Disk L2 Limits
Code L2 Limits
Arch
Impl
Arch
Impl
Tested
Max LV size
(64TB) 16TB
284
284
16EB
264
256TB
Max PV size
241 (2TB)
264
264
264
264
16TB
Max Volume
Groups (VG)
256
N/A
N/A
211
2304*
Max Logical
Volumes (LV)
255
232-1
232-1
211-1
2047
Max Physical
Volumes (PV)
255
232-1
232 -1
232-1
232-1
2047
Max number
of mirrors
3
254
254
7
6
6
Max Nodes
for mirroring
2
N/A
N/A
N/A
16
16
216 (64K)
264
264
264
264
225
1MB – 256MB
1MB –
4TB
1MB–4TB
1MB–
4GB
1MB–
4GB
1MB –
256MB
28 way (255)
232 -1
way
232-1
232-1
232-1
511
Max number
of Extents/VG
Max extent
size (range)
Max stripe
width
224-1
LVM and VxVM Features
Feature
HP-UX 11.31
SAN Boot Support
RAID0
RAID1
Mirrored Stripes (RAID1+0)
Striped Mirrors (RAID 0+1)
Max Volume Size
Max Vol/Volume Group
Max Volume Groups
Rootability
Online Volume Reconfiguration
Fast Resync
Cluster Aware
Static Path Failover (A/P)
Dynamic Path Load balancing (A/A)
# Mirrors
Dynamic Root Disk (DRD)
HP Storage Device Support
3rd Party Storage Support
LVM, MD/UX
VxVM 4.1/5.0 1
Y
Y
MD/UX
MD/UX
MD/UX
Y3
Y
FULL license4
FULL license
FULL license
16EB (256TB)2
2048 (511) 2
2304 (768) 2
Y
SLVM
32TB
Unlimited
Unlimited
Y
FULL license5 + CVM5
MD/UX
Y(SLVM)
Y
Y
6
Y
ALL
Limited6
FULL license + VVR
Y(CVM)
Y
FULL license
32
Y3
Limited6
Limited6
1 – VxVM 4.1 available with 11.31, VxVM 5.0 on 11.31
available H2CY07
3 - VxVM 5.0 only
4 - Root Mirroring part of BASE
6 – Validate device in question with HP or Symantec qualification list:
http://www.hp.com/products1/serverconnectivity/mass_storage_devices.html
2 – supported limit
5 – For all features
http://ftp.support.veritas.com/pub/support/products/VolumeManager_UNIX/277390.pdf
LVM and VxVM Features - continued
Feature
LVM, MD/UX
Instant Snapshots
Portable Data Containers
Multi-volume File System support
Dynamic Storage Tiering support
Dynamic LUN expansion
Extended Campus, Metrocluster,
and Continental cluster
integration
Disk group configuration from
any node
Number Of Nodes
HP-UX 11.31
VxVM 4.1/5.0
N
N
N
N
Y
Y
FULL license
FULL license1
FULL license
FULL license
Y
Y
Y
N
16
4 – VxVM & CVM Base
16/8 – FULL License VxVM/CVM
Storage Distance Support
(standard/Metrocluster)
Rolling Upgrade
RAID 5
Cluster lock support
Hot Relocation
Heterogeneous platforms
1 – VxVM 5.0 Only
2 – AIX and Solaris have similar cluster solution
100 km /
300 km
Y
N
Y (disk, lun)
N
Same
commands on
AIX, Linux
100 km /
300 km
Y3
FULL License
N
Y
AIX, Solaris, Linux, Windows
3 – As long as no functional or protocol changes (y within 3.x, n between 3.x = 4.x)
4 – Not supported for all configurations
Scalability: Expanded file system
Supported
File systems
maximums
N/A = Not architecturally limited
HP-UX 11i v1
HP-UX 11i v2
HP-UX 11i v3
N/A
16TB
50TB
VxFS 5.0 - Number of files
VxFS5.0 - File size
VxFS 5.0 - File system size
VxFS 5.0 - # Access Control
List entries
1024
N/A
2TB
32TB
VxFS 4.1 - Number of files
VxFS 4.1 - File size
VxFS 4.1 - File system size
N/A
16TB
32TB
VxFS 3.5 - number of files
N/A
N/A
VxFS 3.5 - File size
2TB
2TB
2TB
32TB
VxVM volume
32TB
32TB
32 TB
VxVM volumes
2^23
2^23
2^23
32
32
32
VxFS 3.5 - File system size
VxVM mirrors
Disk Space Management Tool
Comparison
Whole Disk
LVM
VxVM
10.x,11.x,11i
10.x,11.x,11i
11i
Yes
sam, smh
Yes
sam, smh
Yes
vea
Similar
Similar
Yes
Can partitions span disks?
No
Yes
Yes
Online resizing supported?
Online backups supported?
No
No
Yes
Yes*
Yes
Yes*
Striping (RAID 0) supported?
No
Yes
Yes
Mirroring (RAID 1) supported?
No
Yes*
Yes*
RAID 0+1 and 1+0 supported?
No
Yes*
Yes*
RAID 5 supported?
No
No
Yes*
Dynamic relayout support?
Dynamic multi-pathing support?
No
No
HP-UX versions supported
Boot disk support?
GUI configuration tool available?
Available on other platforms?
No
Yes*
Active/Passive Active/Active*
* Features indicated by an asterisk may require an extra license
Benefits of HP-UX Native Multipathing
•
•
•
•
•
•
•
Zero-Config. No initial setup or configuration needed. No
license or add-on software installation needed.
Free. Comes bundled with Core HP-UX. Is not an add-on
product. No additional license fee for Native MP.
Transport Aware Failover/Failback.
− Fibre Channel RSCN based failover/failback.
Detection and failover of all paths leading to affected
component (hba, target port, switch, inter-switch link).
Fully parallel path failover/failback and path monitoring.
Built into Storage Stack. More responsive and better
performing.
Exploits both server platform and storage characteristics.
Benefits of HP-UX Native Multipathing continued
•
•
•
•
Tested as a core part of HP-UX. Engineered for
scalability, reliability, performance, serviceability.
Time to Market support for new technologies. Being a
core part of HP-UX, new technologies are supported when
released (ex : HPVM, VPARs, VSE, new transports like
SAS, storage management products, etc).
Simplifies Storage Management. Eliminates an entire
layer in the IO Stack (add-on MP product) and reduces
amount of management overhead for sysadmin. No more
need to install, tune, update, monitor add-on MP product.
Ideal for Oracle ASM. Oracle ASM provides its own
volume management capability and Native MP is a good fit
in that environment since the customer does not need to
install VxVM solely for DMP capabilities.
HP-UX Native Multipathing vs. Classic
MP
•
•
•
•
•
•
•
•
•
Classic MP Product (DMP)
Add-on product layered above IO
stack.
Agnostic of low level SAN and SCSI
events.
Unaware of transport specific
failures (HBA, switch, switch port
offline).
Discovers failures on IO errors.
Needs to issue test IO on IO error.
Path Failover initiated following IO
error and Test IO failure.
Each such path failure detected
individually on IO error on each path.
Need to ping each failed path to
detect recovery.
•
•
•
•
•
•
•
HP-UX 11iv3 Native MP
Not an add-on product. Built into
the IO stack and HBA drivers are
multi-path aware.
Aware of SAN and SCSI events
and takes advantage of those.
Transport Aware. (Knows scope of
outage such as HBA, switch, switch
port, tgt port).
Can failover at the scope of outage.
(All paths under HBA, All paths thru
the switch, all paths thru the switch
port, all paths to the tgt port).
SAN Fabric Events signal recovery.
Pinging is avoided.
Fully parallel failover/failback.
Product-use application
Volume Management
Logical Volume Manager (LVM)
If you need:
•
No-cost solution
• Basic Volume
Management-default
Operating Environment
install
• Simple operation
• SAN boot
• Disk-mirroring capability
with MirrorDisk/UX
VxVM
If you need:
•
Advanced Volume
Management solution
• Instant Snapshots
• Cross-platform data
sharing
• QoSS support
• Raid 5
• Hot Relocate/Unrelocate
• CFS support
Product-use application
File Systems
VxFS Lite
If you need:
•
•
•
•
•
Basic file system
functionality
A no cost solution
Simple operation
Fast file system
recovery
Direct I/O
performance*
VxFS Full (OnlineJFS) Cluster File System (CFS)
If you need:
If you need:
•
•
•
•
•
•
Online
defragmentation
Online log and file
system resizing
Oracle Database
Management
Concurrent I/O
performance*
File change log
Storage
Checkpoints
•
•
•
•
Functionality for
a clustered
environment
Cluster Volume
Manager
Dynamic Storage
Tiers
Flashsnap
technology
Oracle ASM vs. Classical Storage Layers
Oracle Traditional Options - Raw versus
Cooked and With or Without an LVM
April 7, 2016
Disk Groups in the ASM Architecture
Oracle ASM Basics
•
An ASM file system layer is implicitly created within a disk
group. ASM provides a vertical integration of the file
system and volume manager for Oracle database files.
•
This file system is transparent to users and only
accessible through ASM instance, interfacing databases,
and ASM’s utilities. For example, database backups of
ASM files can be performed only with RMAN.
•
Does not completely bypass the O/S I/O stack. It uses the
asyncdsk driver in order to perform asynchronous I/O.
Without it, the logwriter and dbwriters would be doing
direct, synchronous I/O and would not be able to perform
acceptably in high IO workloads. The I/O requests also
have to pass though the SCSI and FC layers.
Oracle ASM Basics - continued
The functionality of an ASM instance can be
summarized as follows:
•
Manages groups of disks, called disk groups.
•
Protects the data within a disk group. *
•
Provides near-optimal I/O balancing without any
manual tuning.
•
Enables the user to manage database objects such
as tablespaces without needing to specify and track
filenames.
•
Supports large files.
Oracle ASM Basics - continued
•
One can use Oracle Enterprise Manager (EM) or
the Database Configuration Assistant (DBCA) for
a GUI interface to Automatic Storage
Management that replaces the use of SQL or
SQL*Plus for configuring and altering disk groups
and their metadata.
•
DBCA eases the configuring and creation of the
database.
•
EM provides an integrated approach for
managing both the ASM instance and database
instance.
Oracle ASM Limits
•
63 disk groups in a storage system.
•
10,000 ASM disks in a storage system.
•
4 Petabyte maximum storage for each ASM disk.
•
40 Exabyte maximum storage for each storage
system.
•
140 Petabyte in external redundancy (no ASM
mirroring).
•
42 Petabyte in normal redundancy (2-way ASM
mirroring).
•
15 Petabyte in high-redundancy (3-way ASM
mirroring).
Oracle ASM Limits
•
ASM disk group that is implemented with External
Redundancy has a maximum file size of 35 TB.
•
ASM Disk Group that is implemented with Normal
Redundancy has a maximum file size of 5.8 TB.
•
ASM Disk Group that is implemented with High
Redundancy has a maximum file size of 3.9 TB.
•
1 million files for each disk group.
•
2.4 Terabyte maximum storage for each file.
File Types Supported by Automatic
Storage Management
File
Supported Default
Templates
Control files
yes
CONTROLFILE
Datafiles
yes
DATAFILE
Redo log files
yes
ONLINELOG
Archive log files
yes
ARCHIVELOG
Trace files
no
N/A
Temporary files
yes
Datafile backup pieces
yes
Datafile incremental backup pieces
BACKUPSET
yes
Archive log backup piece
Datafile copy
TEMPFILE
yes
yes
BACKUPSET
BACKUPSET
DATAFILE
Persistent initialization parameter file (SPFILE) yes
Disaster recovery configurations
PARAMETERFILE
yes
DATAGUARDCONFIG
Flashback logs
yes
FLASHBACK
Change tracking file
yes
CHANGETRACKING
Data Pump dumpset
yes
DUMPSET
Automatically generated control file backup
yes
AUTOBACKUP
Cross-platform transportable datafiles
yes
XTRANSPORT
O/S files
April 7, 2016
no
N/A
Oracle ASM - Pros
•
ASM provides some file system and volume
management capabilities for Oracle database files
only. These include DB control files, redo logs,
archived redo logs, data files, spfiles and Oracle
Recovery Manager (RMAN) backup files (see
previous slide).
•
File-level striping/mirroring.
•
Ease of manageability. Instead of running LVM
software, run an ASM instance, a new type of
"instance" that largely consists of processes and
memory and stores its information in the ASM disks it
is managing.
•
Attempts to identify the configuration errors. *
Oracle ASM - Pros
•
Gives Oracle Corporation control over the
storage system, which makes them happy, so
they promote it heavily. *
•
No large Unix-level administration needed.
•
Provides a single point of support (Oracle) so
there is no “finger-pointing”.
•
Provides easy management of block devices (raw
partitions).
•
Automatically moves hot blocks to the outside of
the disk.
•
Vendor and operating system neutral.
Oracle ASM - Pros
•
Included in the Oracle license so no additional cost
for the software or its support.
•
Supports very large disk groups and datafiles.
•
Database File System with performance of RAW
I/O.
•
Supports clustering (RAC) and single instance.
•
Automatic data distribution.
•
Memory requirements for an ASM instance are
small. 100 MB of RAM is typically all that is
required to run an ASM instance on a production
server.
Oracle ASM - Pros
•
On-line add/drop/resize disk with minimum data
relocation.
•
Automatic file management. *
•
Flexible mirror protection.
•
Inode locks not applicable to ASM.
•
Ability to grow diskgroup capacity on the fly.
•
Supports multiple database instances running on
a single host, and does not have its own data
dictionary.
Oracle ASM – Pros (Oracle 11gr)
•
Fast mirror resynchronization. *
•
Preferred mirror read in a cluster. **
•
Support for large ALU. ***
•
Variable size extents. ****
•
Rolling upgrade and patching.
•
Table level migration wizard in EM.
•
New ASMCMD commands.
•
New SYSADM privilege – separate from the
SYSDNA privilege.
•
More flexible FORCE options to MOUNT or DROP
disk groups.
Oracle ASM - Cons
•
ASM cannot be used for Oracle executables and nondatabase files.
•
ASM files can only be managed through an Oracle
application such as RMAN. This can be a weakness if you
prefer third-party backup software or simple backup
scripts. Cannot store CRS files or database software.
Cannot manage ASM through standard Unix tools.
•
Potentially disrupts the balance of power between the
Unix Systems Administration groups, and the
Database/DBA groups. Traditionally the former group
manages disks, hardware, and the operating system
level, leaving the DBAs to coordinate with them for new
resources. This would change that balance, which could
cause some resistance. *
Oracle ASM - Cons
•
ASM does not have multi-pathing capability. It assumes the
underlying O/S will provide this functionality. In HP-UX,
multi-pathing is provided by a Volume Manager feature
such as PVLinks in the HP-UX Logical Volume Manager
(LVM), native multipathing in HP-UX 11.31, or DMP in
Veritas Volume Manager from Symantec (VxVM), or by
other third-party software such as Securepath or
Powerpath.
•
ASM is still in the enterprise computing, relatively new.
There are a number of vendors whose core business has
been in the logical volume manager/file system space for
years. Often, maturity matters a lot when it comes to
software systems, reliability, and proven success rates.
•
New technology to get familiar with.
Oracle ASM - Cons
•
Automatic Storage Management load balances file
activity by uniformly distributing file extents across all
disks in a disk group. For this technique to be effective
it is important that the disks used in a disk group be of
similar performance characteristics.
There may be situations where it is acceptable to
temporarily have disks of different sizes and
performance co-existing in a disk group (for example,
when migrating from an old set of disks to a new set of
disks). The new disks would be added and the old
disks dropped. As the old disks are dropped, their
storage is migrated to the new disks while the disk
group in online.
Oracle ASM - Cons
•
There is no shared awareness of LUN use between ASM
and LVM or VxVM. It means that the system administrator
must be careful not to accidentally allocate a LUN already
allocated for LVM or VxVM use to ASM use (or vice-versa).
•
ASM is not an enterprise-class file system.
•
ASM is a proprietary solution.
•
The customer is dependant on the reliability of the new
ASM code stack. *
•
Does not offer network monitoring.
•
Be careful about ASM hidden parameters.
Oracle ASM - Cons
•
Not for high I/O environments (that is what some
tests claim).
•
Everything is single threaded through one process
at a very low level. *
•
If one uses Oracle ASM and CRS, they will still
require a 3rd party clustering solution to support the
non-Oracle data. They will then have to manage
multiple clustering solutions.
Oracle ASM - Cons
•
DBAs must still watch and then perform the task of adding and
removing disks to an ASM disk group when needed. This leads
back to the problem of DBAs under allocating, and worse yet,
over allocating disk storage, just to be safe, which recreates the
problem of wasted space and leads to higher than needed
storage costs.
This is where thin provisioning comes into play. Thin provisioning
will automatically allocate on a just-enough and just-in-time basis
which relieves the DBA from both having to watch and then add
or remove disk to a disk group. Oracle's ASM and thin
provisioning could be combined to offer a complete, end-to-end,
storage solution. Oracle's ASM feature would create, allocate,
place, and rebalance data files for performance and thin
provisioning would dedicate disk space on the fly and only when
needed.
Oracle ASM - Cons
•
Configuration details and performance metrics are exposed via
V$ views. Other possibilities are the command line interface,
asmcmd and the graphical interface of OEM.
•
Metadata are however partially hidden to the end user. That is
the mapping between physical storage, ASM allocation units, and
database files is not completely exposed via V$ views. It was
found that is possible to query such information via
undocumented X$ tables. For example, it is possible to
determine the exact physical location on disk of each extent (or
mirror copies of extents) for each file allocated on ASM (and if
needed access the data directly via the O/S). This can be used
by Oracle DBAs wanting to extend their knowledge of the inner
workings of the ASM or wanting to diagnose hotspots and ASM
rebalancing issues.
What about Non-Database Files?
•
Since ASM supports only database data files and log files,
the following storage management methods are required
for non-database files if ASM is used:
•
A) A local file system or Cluster File System for Oracle
Clusterware binaries and configuration files and RAC
binaries and configuration files.
•
B) Shared storage for Oracle Clusterware data: voting
disk and OCR files. This storage has to be configured
either as shared raw devices, shared raw volumes, or files
in the Cluster File System. Oracle Clusterware needs to
be up and running before the ASM daemon can start.
Therefore, ASM cannot be used for Oracle Clusterware
data.
Some of the Oracle Plans and Efforts *
•
Elimination of Symantec by destroying Veritas Cluster as a
viable product (replaced with CRS), and the same for Veritas
File System (replaced by ASM).
•
Elimination of Sun by adopting Linux as a low-cost alternative
to Solaris (Sun has now been acquired by Oracle who has
been primarily responsible for its downfall).
•
Attempted elimination of Red Hat by “migrating” Red Hat Linux
under Oracle Enterprise Linux (limited success so far).
•
Challenge to EMC and NetApp in the area of storage using the
Oracle Exadata Storage (version 1 was with HP and justannounced version 2 on Intel X86 and Linux).
•
Changes in licensing structure.
•
Acquisition of Sun Microsystems.
Capability compare
Business Benefit
Vendor
Product(s)
HP
OnlineJFS+LVM/Mir
rorDisk/UX (Single
Instance only)
HP
OnlineJFS+VxVM/O
DM Storage
Management for
Oracle (Single
Instance only)
Cluster file
system(both RAC &
Single Instance)
No file system –
raw volumes
(RAC ONLY)
Oracle
Automatic Storage
Management (RAC
& SI)
HP
HP
Oracle
executable and
non-database Multi-pathing
files
capability
Strong I/O
fencing
Single-instance
Near raw
Database RAC Database
database
object storage object storage performance
Yes
Yes
Yes
Yes
no
Yes
Yes
Yes
Yes
Yes
no
Yes
Yes
Yes
Yes
y
Yes
y
No
No
Yes
No*
No
Yes
YES – it is raw
volume
No
No
(limited)
No
Yes
Yes
Capability compare - continued
Business Benefit
Vendor
HP
HP
HP
HP
Oracle
Product(s)
OnlineJFS+LVM/Mirr
orDisk/UX (Single
Instance only)
OnlineJFS+VxVM/OD
M (Storage
Management for
Oracle)(Single
Instance only)
Cluster file
system(both RAC &
Single Instance)
No file system – raw
volumes
(RAC ONLY)
Automatic Storage
Management (RAC &
SI)
Ability to use
Ease of
standard UNIX
management
commands to Proven, timeAll disaster(e.g., backup
move or copy tested quality recovery needs and database
data
and reliability
can be met
maintenance)
Yes
Yes
Yes
y
Yes
Yes
Yes
y
y
y
n
y
y
n
No
No
No
y
Oracle ASM versus HP-UX SLVM **
• ASM lacks Multipathing: The ASM-over-SLVM configuration
provides multipathing for ASM disk groups (using LVM PV
Links or storage based multipathing). This is not an issue on
HP-UX 11i v3 (native multipathing).
•
ASM-over-SLVM enables the HP-UX devices used for disk
group members to have the same names on all nodes, easing
ASM configuration. Generally, when using raw disk devices
with ASM, the device files are created under /dev/oracle, so
they can be the same on all nodes.
•
ASM-over-SLVM protects ASM data against inadvertent
overwrites from nodes inside/outside the cluster. If the ASM
disk group members are raw disks, there is no protection
currently against these disks being setup in LVM or VxVM
volume/disk groups. *
•
There are some disadvantages too. **
Oracle Discontinues Raw/Block Device
Support
In release 11.2, the Oracle Installer and DBCA
(Database Configuration Assistant) will no longer
support raw/block devices for database files. In
addition, there will no longer be raw/block device
support for storing the OCR and Voting Disks for new
installs. Customers who create a new 11.2 database
will need to store their database files in either ASM, a
file system, or on an NFS filer. RAC database files
must
be stored on ASM, a certified clustered file system, or
a
Oracle ASM on HP-UX – Some
Experiences
•
It was found Oracle ASM was a good choice to simplify
Oracle related storage and volume management
− However do not use ASM redundancy features
•
The HP-UX 11v3 MSS (Mass Storage Stack) and Oracle
ASM were a “killer combo”.
•
ASM works very well for Oracle data management, but a
severe performance penalty gets incurred if the ASM
redundancy features are used on top of the storage array
redundancy features.
One of HP-UX Solutions
•
HP-UX file systems can provide read/write
performance 95-98% as fast as a raw-device
setup.*
•
OJFS 5.0.1: performance impact virtually
eliminated.
•
The manageability and reduced administration
from a file system, with near-raw performance.
* Currently this is offered through the purchase of a Storage Management Suite bundle
4
HP-UX 11iv3 Mass Storage Stack
•
•
Transparent native multipathing to LUNs is useful
and necessary – decreases work and rework.
Load balancing increased performance
− “least command load” policy provided the best
performance for this BI workload
− “round robin” policy, the default, provided slightly less
performance
•
Persistent LUN bindings and LUN DSF’s reduce
design and maintenance efforts significantly for
mid-range and high-end BI implementations
containing large storage subsystems.
Some Best Practices for Oracle ASM
with HP StorageWorks
As a result of testing, a set of best practices for using Oracle ASM
with HP StorageWorks on HP-UX servers is presented:
• Configure ASM disk groups to use external redundancy.
• When building a disk group or adding to an existing one, use
disks of similar capacity and performance characteristics in the
same disk group.
• To leverage I/O distribution across as many resources as
possible, it is best to present more than one LUN to a disk group
(allowing ASM to do the striping).
• Use HP Secure Path with ASM on HP-UX 11.23 or MSS (Mass
Storage Stack) on HP-UX 11.31 because it complements the
high availability and performance of the entire stack.
Some Best Practices for Oracle ASM
with HP StorageWorks - continued
•
Each device (LUN) should be managed either by
Oracle ASM or by HP-UX LVM, but not both.
• Care should be taken not to attempt inadvertently
to manage an ASM disk by a traditional volume
manager or vice versa.
• Configure async I/O (Please consult the Oracle
Administration Guide documentation).
• Use "insf" instead of "insf -e" to create the
devices associated with new hardware because
"insf -e" will reset the ownership to user "bin“.