Oracle ACFS Support on Oracle Exadata Database Machine (Linux only) (Doc ID 1929629.1).pdf

Oracle ACFS Support on Oracle Exadata Database Machine (Linux only) (Doc ID
1929629.1)
In this Document
Purpose
Details
Oracle ACFS Support on Exadata Database Machine
Requirements and Best Practice Recommendations
Oracle ACFS Use Cases on Exadata Database Machine
References
APPLIES TO:
Oracle Exadata Storage Server Software
Linux x86-64
PURPOSE
The purpose of this document is to explain Oracle Automatic Storage Management File System (ACFS) support on
Oracle Exadata Database Machine.
DETAILS
Oracle ACFS Support on Exadata Database Machine
Starting with Oracle Grid Infrastructure 12c Release 12.1.0.2, Oracle Automatic Storage Management File System
(ACFS) supports all database files and general purpose files on Oracle Exadata Database Machine running Oracle Linux
on database servers. Please refer to the Oracle Automatic Storage Management Administrator's Guide for details on
how to configure and administer Oracle ACFS.
On Exadata Database Machine, Oracle ACFS supports the following database versions:
Database 11.2.0.4 and higher
Database 10.2.0.4 and 10.2.0.5 - see Document 1965897.1 for additional details
Review Document 888828.1 for specific Grid Infrastructure version and patch requirements.
The following list provides important information about using Oracle ACFS with Exadata:
Oracle ACFS supports all database file types, database software binaries (database ORACLE_HOME), and general
purpose files.
IMPORTANT: Mission-critical databases should not use ACFS for any of the following:
Database files (e.g. data files, control files, and redo log files)
Database home software/binaries (ORACLE_HOME)
Administrative files (e.g. diagnostic, trace, and audit directories - audit_file_dest,
background_dump_dest, core_dump_dest, diagnostic_dest, user_dump_dest)
Store database files in ASM. Use a local disk file system (e.g. /u01) to store database home software/binaries
and administrative files.
Oracle ACFS snapshots should only be used for test/development databases.
Document 1929629.1 https://mosemp.us.oracle.com/epmos/faces/DocumentDispla...
第1页 共3页2022/10/11 18:41
Do not colocate test/development and production in the same VM/cluster.
All Oracle ACFS advanced functionality is supported on Exadata. This includes:
Snapshots, replication, tagging, encryption, and HANFS.
Database files stored in ACFS on Exadata storage are subject to the following guidelines and restrictions:
Oracle ACFS replication or security/encryption/audit is only supported with general purpose files. These functions
are not supported with database files in an Oracle ACFS file system.
Oracle ACFS does not support the following Exadata features: Smart Scan, Smart Flash Logging, Storage
Indexes, Network Resource Management, Hardware Assisted Resilient Data (HARD) checks.
Exadata Smart Flash Cache will cache both read and write operations for database files stored in ACFS. Read
operation caching has no minimum software requirement. Write operation caching requires the following
software:
On storage servers, one of the following:
Exadata 20.1.1 or higher
Exadata 19.3.11 or higher 19.3
Grid Infrastructure 19c or higher
Grid Infrastructure 19c requires the Patch 31648120 to be installed in the GI home.
Oracle ACFS supports IORM tagging of database files.
Hybrid Columnar Compression (HCC) support requires Patch 19136936.
No specialized cache hints are passed from the Database to the Exadata Storage layer, which means the Smart
Flash Cache heuristics are based on I/O size, similar to any other block storage caching technology.
Requirements and Best Practice Recommendations
Mission-critical databases should not use ACFS for database files (e.g. data files, control files, and redo log files),
database home software/binaries (ORACLE_HOME), and administrative files (e.g. diagnostic, trace, and audit
directories). Use ASM directly (not ACFS) to store database files. Use a local disk file system (e.g. /u01) to store
database home software/binaries and administrative files.
To obtain optimal overall performance and high availability, Grid Infrastructure of 19.10 or later with the
following additional fixes should be used. This version/patch combination includes a number of stability and
performance fixes for ACFS. For all other uses, please review Document 888828.1 for version recommendations.
Bug 31897853
Bug 32492090
Bug 32677159
Bug 31996407
Bug 31957860
Bug 31787103
When creating an ASM disk group for ACFS use:
Create ACFS volumes in a disk group separate from where database files are stored in ASM (i.e. separate
from DATA and RECO).
For each disk group that contains an Oracle ADVM volume that is intended to hold database files, set the
compatible.asm and compatible.advm disk group compatibility attributes to 12.1.0.2, or higher. The
recommendation is to set compatible.asm and compatible.advm equal to the current ASM version. See
the Oracle Automatic Storage Management Administrator's Guide for details.
When creating ACFS volumes:
Create ACFS volumes in a disk group separate from where database files are stored in ASM (e.g. separate
from DATA and RECO).
Volumes created while running 12.1.0.2 or higher already default to the high performance configuration
(stripe columns = 8 and stripe width = 1MB).
For volumes created before 12.1.0.2, set the stripe columns to 1, or set the stripe columns to 8 and the
stripe width to 1MB. See the Oracle Automatic Storage Management Administrator's Guide for details.
When creating ACFS file systems:
Create ACFS file systems with metadata block size of 4096 (e.g. mkfs -t acfs -i 4096).
When storing database files in ACFS:
Set the FILESYSTEMIO_OPTIONS initialization parameter to "setall". Other settings are not supported.
Use a 4 K or larger database block size and tablespace block size with Oracle ACFS for best performance.
If a data file is configured to automatically extend, then the size of the increments should be large
enough to ensure that the extend operation occurs infrequently. Frequent automatic extends have a
Document 1929629.1 https://mosemp.us.oracle.com/epmos/faces/DocumentDispla...
第2页 共3页2022/10/11 18:41
negative performance impact.
When using ACFS snapshots:
To obtain optimal database performance with snapshots, the snapshots must be created after the ADVM
compatibility attribute is set to 12.1 or higher, however recommended it be set to match the version of GI
installed.
Running a workload in a snapshot reduces resources for the primary workload running on the base files
because the storage is shared between the base file system and the snapshots. To run test scenarios in
Oracle ACFS snapshots without impacting the primary workload, copy the file system and then run test
workloads on snapshots created in the copied file system.
Oracle ACFS Use Cases on Exadata Database Machine
The initial design objective of ACFS on Exadata was to provide a general purpose, industry standard, POSIX
compliant cluster file system on Exadata platforms to manage all general purpose files with good performance.
Oracle ACFS also supports the industry standard network protocols NFS/CIFS and offers a highly available NFS
solution with the HANFS feature.
Oracle ACFS supports Oracle Database 10g 10.2.0.4 and 10.2.0.5 releases. Customers can migrate their 10g
Oracle Databases to Exadata and upgrade later. This makes Exadata a more flexible platform enabling
consolidation and easy migration for customers. See Document 1965897.1 for additional details.
Oracle ACFS may be configured as a shared Oracle database home, middleware and application homes/binaries
and shared files like Oracle Golden Gate trail files.
Oracle ACFS can be configured for use with the database particularly to leverage Oracle ACFS snapshots for
database testing and development.
REFERENCES