corrections around ASM recommendations

msjuergent · msjuergent · commit 338777481776 · 2024-04-20T06:19:34.000-07:00
diff --git a/articles/sap/workloads/dbms-guide-oracle.md b/articles/sap/workloads/dbms-guide-oracle.md
@@ -51,16 +51,16 @@ Installing or migrating existing SAP on Oracle systems to Azure, the following d
 
 1.  Use the most [recent Oracle Linux](https://docs.oracle.com/en/operating-systems/oracle-linux/8/) version available (Oracle Linux 8.6 or higher)
 2.  Use the most recent Oracle Database version available with the latest SAP Bundle Patch (SBP) (Oracle 19 Patch 15 or higher) [2799920 - Patches for 19c: Database](https://launchpad.support.sap.com/#/notes/2799920)
-3.  Use Automatic Storage Management (ASM) for small, medium and large sized databases on block storage
+3.  Use Automatic Storage Management (ASM) for small, medium, and large sized databases on block storage
 4.  Azure Premium Storage SSD should be used. Don't use Standard or other storage types.
 5.  ASM removes the requirement for Mirror Log. Follow the guidance from Oracle in Note [888626 - Redo log layout for high-end systems](https://launchpad.support.sap.com/#/notes/888626)
 6.  Use ASMLib and don't use udev
 7.  Azure NetApp Files deployments should use Oracle dNFS (Oracle’s own high performance Direct NFS solution)
-8.  Large Oracle databases benefit greatly from large SGA sizes. Large customers should deploy on Azure M-series with 4 TB or more RAM size.
+8.  Large Oracle databases benefit greatly from large System Global Area (SGA) sizes. Large customers should deploy on Azure M-series with 4 TB or more RAM size.
     - Set Linux Huge Pages to 75% of Physical RAM size
-    - Set SGA to 90% of Huge Page size
+    - Set System Global Area (SGA) to 90% of Huge Page size
     - Set the Oracle parameter USE_LARGE_PAGES = **ONLY** - The value ONLY is preferred over the value TRUE as the value ONLY is suppossed to deliver more consistent and predictable performance. The value TRUE may allocate both large 2MB and standard 4K pages. The value ONLY is going to always force large 2MB pages. If the number of available huge pages is not sufficient or not correctly configured, the database instance is going to fail to start with error code: *ora-27102 :  out of memory Linux_x86_64 Error 12 : cannot allocate memory*. If there is insufficient contiguous memory Oracle the Operating System may need to be restarted and/or the Operating System Huge Page parameters reconfigured
-9.  Oracle Home should be located outside of the “root” volume or disk. Use a separate disk or ANF volume. The disk holding the Oracle Home should be 64GB or larger
+9.  Oracle Home should be located outside of the "root" volume or disk. Use a separate disk or ANF volume. The disk holding the Oracle Home should be 64GB or larger
 10. The size of the boot disk for large high performance Oracle database servers is important. As a minimum a P10 disk should be used for M-series or E-series. Don't use small disks such as P4 or P6. A small disk can cause performance issues.
 11. Accelerated Networking must be enabled on all VMs. Upgrade to the latest OL release if there are any problems enabling Accelerated Networking
 12. Check for updates in this documentation and SAP note [2039619 - SAP Applications on Microsoft Azure using the Oracle Database: Supported Products and Versions - SAP ONE Support Launchpad](https://launchpad.support.sap.com/#/notes/2039619)
@@ -77,7 +77,7 @@ There are two recommended storage deployment patterns for SAP on Oracle on Azure
 1.  Oracle Automatic Storage Management (ASM)
 2.  Azure NetApp Files (ANF) with Oracle dNFS (Direct NFS)
 
-Customers currently running Oracle databases on EXT4 or XFS file systems with LVM are encouraged to move to ASM. There are considerable performance, administration and reliability advantages to running on ASM compared to LVM. ASM reduces complexity, improves supportability and makes administration tasks simpler. This documentation contains links for Oracle DBAs to learn how to install and manage ASM.
+Customers currently running Oracle databases on EXT4 or XFS file systems with Logical Volume Manager (LVM) are encouraged to move to ASM. There are considerable performance, administration and reliability advantages to running on ASM compared to LVM. ASM reduces complexity, improves supportability and makes administration tasks simpler. This documentation contains links for Oracle Database Administrators (DBAs) to learn how to install and manage ASM.
 
 Azure provides [multiple storage solutions](../../virtual-machines/disks-types.md).  The table below details the support status 
 
@@ -103,18 +103,18 @@ Additional considerations that apply list like:
    - LVM - If disks with 4K Native geometry are used, FILESYSTEMIO_OPTIONS = ASYNC 
 2. Oracle 19c and higher fully supports 4K Native sector size with both ASM and LVM
 3. Oracle 19c and higher on Linux – when moving from 512e storage to 4K Native storage Log sector sizes must be changed  
-4. To migrate from 512/512e sector size to 4K Native Review (Doc ID 1133713.1) – see section “Offline Migration to 4Kb Sector Disks”
-5.	SAPInst will write to the pfile during installation.  If the $ORACLE_HOME/dbs is on a 4K disk set filesystemio_options=asynch and see the Section “Datafile Support of 4kb Sector Disks” in MOS Supporting 4K Sector Disks (Doc ID 1133713.1)
+4. To migrate from 512/512e sector size to 4K Native Review (Doc ID 1133713.1) – see section "Offline Migration to 4Kb Sector Disks"
+5.	SAPInst writes to the pfile during installation.  If the $ORACLE_HOME/dbs is on a 4K disk set filesystemio_options=asynch and see the Section "Datafile Support of 4kb Sector Disks" in MOS Supporting 4K Sector Disks (Doc ID 1133713.1)
 5. No support for ASM on Windows platforms
-6. No support for 4K Native sector size for Log volume on Windows platforms.  SSDv2 and Ultra Disk must be changed to 512e via the “Edit Disk” pencil icon in the Azure Portal
+6. No support for 4K Native sector size for Log volume on Windows platforms.  SSDv2 and Ultra Disk must be changed to 512e via the "Edit Disk" pencil icon in the Azure Portal
 7. 4K Native sector size is supported only on Data volumes for Windows platforms.  4K isn't supported for Log volumes on Windows
-8. It's recommended to review these MOS articles:
+8. We recommended to review these MOS articles:
     - Oracle Linux: File System's Buffer Cache versus Direct I/O (Doc ID 462072.1)
     - Supporting 4K Sector Disks (Doc ID 1133713.1)
     - Using 4k Redo Logs on Flash, 4k-Disk and SSD-based Storage (Doc ID 1681266.1)
     - Things To Consider For Setting filesystemio_options And disk_asynch_io (Doc ID 1987437.1)
 
-It's recommended to use Oracle ASM on Linux with ASMLib.  Performance, administration, support and configuration are optimized with deployment pattern.  Oracle ASM and Oracle dNFS are going to  set the correct parameters or bypass parameters (such as FILESYSTEMIO_OPTIONS) and therefore deliver better performance and reliability. 
+We recommended to use Oracle ASM on Linux with ASMLib.  Performance, administration, support and configuration are optimized with deployment pattern.  Oracle ASM and Oracle dNFS are going to  set the correct parameters or bypass parameters (such as FILESYSTEMIO_OPTIONS) and therefore deliver better performance and reliability. 
 
 
 ### Oracle Automatic Storage Management (ASM)
@@ -129,7 +129,7 @@ Checklist for Oracle Automatic Storage Management:
 5.  ASM Disk Groups configured as per Variant 1, 2 or 3 below
 6.  ASM Allocation Unit size = 4MB (default). VLDB OLAP systems such as BW may benefit from larger ASM Allocation Unit size. Change only after confirming with Oracle support
 7.  ASM Sector Size and Logical Sector Size = default (UDEV isn't recommended but requires 4k)
-8.	If the COMPATIBLE.ASM disk group attribute is set to 11.2 or greater for a disk group, you can create, copy, or move an Oracle ASM SPFILE into ACFS file system.  Review the Oracle documentation on moving pfile into ACFS. SAPInst will not create the pfile in ACFS by default
+8.	If the COMPATIBLE.ASM disk group attribute is set to 11.2 or greater for a disk group, you can create, copy, or move an Oracle ASM SPFILE into ACFS file system.  Review the Oracle documentation on moving pfile into ACFS. SAPInst isn't creating the pfile in ACFS by default
 8.  Appropriate ASM Variant is used. Production systems should use Variant 2 or 3
 
 ### Oracle Automatic Storage Management Disk Groups
@@ -154,7 +154,7 @@ Oracle ASM disk group recommendation:
 |ASM Disk Group Name   |Stores                        | Azure Storage       |
 |----------------------|------------------------------|--------------------|
 | **+DATA**  |All data files |3-6 x P 30 (1 TiB) |
-| |Control file (first copy) | To increase DB size, add extra P30 disks |
+| |Control file (first copy) | To increase database size, add extra P30 disks |
 | |Online redo logs (first copy) |                    |
 | **+ARCH** |Control file (second copy) | 2 x P20 (512 GiB) |
 | |Archived redo logs  |  |
@@ -176,12 +176,12 @@ Major differences to Variant 1 are:
 1.  Separate Oracle ASM Disk Group for each database
 2.  \<DBNAME\>+“\_” is used as a prefix for the name of the DATA disk group
 3.  The number of the DATA disk group is appended if the database spans over more than one DATA disk group
-4.  No online redo logs are located in the “data” disk groups. Instead an extra disk group is used for the first member of each online redo log group.
+4.  No online redo logs are located in the "data" disk groups. Instead an extra disk group is used for the first member of each online redo log group.
 
 | ASM Disk Group Name   | Stores |Azure Storage |
 |---|----|---|
 | **+\<DBNAME\>\_DATA[#]** | All data files | 3-12 x P 30 (1 TiB) |
-|   | All temp files | To increase DB size, add extra P30 disks |
+|   | All temp files | To increase database size, add extra P30 disks |
 |   |Control file (first copy)   |  |
 | **+OLOG**  | Online redo logs (first copy) | 3 x P20 (512 GiB) |
 | **+ARCH**  | Control file (second copy) |3 x P20 (512 GB) |
@@ -201,10 +201,10 @@ Usually customers are using RMAN, Azure Backup for Oracle and/or disk snap techn
 |ASM Disk Group Name | Stores | Azure Storage  |
 |---|---|---|
 | **+\<DBNAME\>\_DATA[#]** | All data files  |5-30 or more x P30 (1 TiB) or P40 (2 TiB) 
-|   | All temp files  To increase DB size, add extra P30 disks |
+|   | All temp files  To increase database size, add extra P30 disks |
 |   |Control file (first copy)  |  |
 | **+OLOG** | Online redo logs (first copy) |3-8 x P20 (512 GiB) or P30 (1 TiB)  |
-|   |    | For more safety “Normal Redundancy” can be selected for this ASM Disk Group |
+|   |    | For more safety "Normal Redundancy" can be selected for this ASM Disk Group |
 |**+ARCH** | Control file (second copy) |3-8 x P20 (512 GiB) or P30 (1 TiB) |
 |   | Archived redo logs | |
 | **+RECO** | Control file (third copy) |3 x P30 (1 TiB), P40 (2 TiB) or P50 (4 TiB) |
@@ -218,7 +218,7 @@ Usually customers are using RMAN, Azure Backup for Oracle and/or disk snap techn
 
 ### Adding Space to ASM + Azure Disks
 
-Oracle ASM Disk Groups can either be extended by adding extra disks or by extending current disks. It's recommended to add extra disks rather than extending existing disks. Review these MOS articles and links MOS Notes 1684112.1 and 2176737.1
+Oracle ASM Disk Groups can either be extended by adding extra disks or by extending current disks. We recommended to add extra disks rather than extending existing disks. Review these MOS articles and links MOS Notes 1684112.1 and 2176737.1
 
 ASM adds a disk to the disk group:
 `asmca -silent -addDisk -diskGroupName DATA -disk '/dev/sdd1'`
@@ -277,7 +277,7 @@ Mirror Log is required on dNFS ANF Production systems.
 
 Even though the ANF is highly redundant, Oracle still requires a mirrored redo-logfile volume. The recommendation is to create two separate volumes and configure origlogA together with mirrlogB and origlogB together with mirrlogA. In this case, you make use of a distributed load balancing of the redo-logfiles.
 
-The mount option “nconnect” is NOT recommended when the dNFS client is configured. dNFS manages the IO channel and makes use of multiple sessions, so this option is obsolete and can cause manifold issues. The dNFS client is going to ignore the mount options and is going to handle the IO directly.
+The mount option "nconnect" is NOT recommended when the dNFS client is configured. dNFS manages the IO channel and makes use of multiple sessions, so this option is obsolete and can cause manifold issues. The dNFS client is going to ignore the mount options and is going to handle the IO directly.
 
 Both NFS versions (v3 and v4.1) with ANF are supported for the Oracle binaries, data- and log-files.
 
@@ -294,7 +294,7 @@ Recommended mount options are:
 
 ### ANF Backup
 
-With ANF, some key features are available like consistent snapshot-based backups, low latency, and remarkably high performance. From version 6 of our AzAcSnap tool [Azure Application Consistent Snapshot tool for ANF](../../azure-netapp-files/azacsnap-get-started.md) Oracle databases can be configured for consistent database snapshots. Also, the option of resizing the volumes on the fly is valued by our customers.
+With ANF, some key features are available like consistent snapshot-based backups, low latency, and remarkably high performance. From version 6 of our AzAcSnap tool [Azure Application Consistent Snapshot tool for ANF](../../azure-netapp-files/azacsnap-get-started.md), Oracle databases can be configured for consistent database snapshots. Also, the option of resizing the volumes on the fly is valued by our customers.
 
 Those snapshots remain on the actual data volume and must be copied away using ANF CRR (Cross Region Replication) [Cross-region replication of ANF](../../azure-netapp-files/cross-region-replication-introduction.md)
 or other backup tools.
@@ -303,7 +303,7 @@ or other backup tools.
 
 ASM is the default recommendation from Oracle for all SAP systems of any size on Azure. Performance, Reliability and Support are better for customers using ASM. Oracle provides documentation and training for DBAs to transition to ASM and every customer who migrated to ASM has been pleased with the benefits. In cases where the Oracle DBA team doesn't follow the recommendation from Oracle, Microsoft and SAP to use ASM the following LVM configuration should be used.
 
-Note that: when creating LVM the “-i” option must be used to evenly distribute data across the number of disks in the LVM group.
+Note that: when creating LVM the "-i" option must be used to evenly distribute data across the number of disks in the LVM group.
 
 Mirror Log is required when running LVM.
 
@@ -386,9 +386,9 @@ Another good Oracle whitepaper [Setting up Oracle 12c Data Guard for SAP Custome
 
 ## Huge Pages & Large Oracle SGA Configurations
 
-VLDB SAP on Oracle on Azure deployments apply SGA sizes in excess of 3TB.  Modern versions of Oracle handle large SGA sizes well and significantly reduce IO.  Review the AWR report and increase the SGA size to reduce read IO. 
+VLDB SAP on Oracle on Azure deployments apply SGA sizes in excess of 3TB. Modern versions of Oracle handle large SGA sizes well and significantly reduce IO. Review the AWR report and increase the SGA size to reduce read IO. 
 
-As general guidance Linux Huge Pages should be configured to approximately 75% of the VM RAM size.  The SGA size can be set to 90% of the Huge Page size.  An approximate example would be a m192ms VM with 4 TB of RAM would have Huge Pages set proximately 3 TB.  The SGA can be set to a value a little less such as 2.95 TB.
+As general guidance Linux Huge Pages should be configured to approximately 75% of the VM RAM size. The SGA size can be set to 90% of the Huge Page size.  An approximate example would be a m192ms VM with 4 TB of RAM would have Huge Pages set proximately 3 TB.  The SGA can be set to a value a little less such as 2.95 TB.
 
 Large SAP customers running on High Memory Azure VMs greatly benefit from HugePages as described in this [article](https://www.carajandb.com/en/blog/2016/7-easy-steps-to-configure-hugepages-for-your-oracle-database-server/)
 
