Exploring Flex ASM and other Oracle features
by Rackspace Technology Staff
Introduction
Oracle© version 12.1.0.1 has incorporated two new features based on Hub and Leaf node architecture.
The new features are:
- Flex Cluster
- Flex Automatic Storage Management (Flex ASM)
Oracle implements the Flex ASM feature by default in Oracle Database Appliance (ODA) version 12.1.x, along with the ASM Dynamic Volume Manager (ADVM) Proxy instance and the Grid Infrastructure Management Repository database, MGMTDB.
This blog explains Flex ASM, ADVM and MGMTDB features and the associated benefits.
Flex ASM
Flex ASM reduces the ASM instance footprint by running on each node of the cluster, and also ensures that the database instance has high availability, in case of ASM instance failure on a particular node. You can configure Flex ASM by using dedicated ASM network, or you can share the private interconnect used by the cluster. The following images show the test cases performed for Flex ASM:
Cluster ASM listener Node1:
Cluster ASM listener Node2:
Use the following code to check the ASM cluster mode and state:
Use the following code to check ASM status:
Use the following code to check ASM configuration:
NOTE: You can reduce an ASM instance count to a minimum of two by using the following command, which shows instances running on Node1
srvctl modify asm -count 2
Then run the following code:
srvctl status asm -node ushodld0001ca
Then run the following code:
srvctl status asm -node ushodld0001ca
Now, you need to stop the ASM instance on Node1 by running the following command:
srvctl stop asm -node ushodld0001ca -stopoption abort –force
Use the following code to check the status:
srvctl status asm
Use the following code to check the cluster:
crsctl check cluster
Use the following code to Verify that all the instances except ASM1 are running on Node1 because of Flex ASM:
Use the following code to verify that instances are running fine on both the nodes even if ASM1 is down:
Use the following code to check the ADVM proxy instance:
Benefits of Flex ASM
Flex ASM has the following benefits:
- Oracle Flex ASM supports larger LUN sizes for Oracle Database 12c clients.
- You can connect to other ASM instance from a different instance (for example, connect to a second +ASM2 instance from the first instance).
- You have the flexibility to rename an ASM Disk in a Disk Group. You can rename the Disk name given by Oracle as needed.
About ADVM proxy instance
The ADVM Proxy is a special Oracle instance, which enables ADVM to connect to Flex ASM and is required to run on the same node as the ADVM and the ASM Cluster File System (ACFS).
The ADVM Proxy instance ORACLE_SID is set to `+APX<node number>`.
MGMTDB - Grid Infrastructure Management Repository database
MGMTDB is a database instance used for storing cluster health monitor data (CHM). In version 11g, this information is stored in berkley database. Starting from version 12.1.0.2, it is part of default (mandatory) installation.
If the existing node crashes. MGMTDB, an Oracle single instance managed by Grid Infrastructure, fails over to the surviving node. You can identify the current master using the following command:
oclumon manage -get MASTER
This instance is managed by using the following SRVCTL commands:
srvctl status mgmtdb
srvctl config mgmtdb
You can see the single listener associated with this database by running the following command:
ps -ef|grep MGMTLSNR.
You can view information about MGMTDB from metalink doc - 1568402.1.
By using the `oclumon` command, you can collect information about the cluster health.
Conclusion
This blog explored Oracle's Flex ASM, the ADVM Proxy instance, and MGMTDB features.
Flex ASM reduces the ASM instance footprint, running on each node of the cluster and also ensures high availability of your database instance, in case the ASM instance fails on a particular node.
The ADVM Proxy is a special Orale instance, which enables ADVM to connect to Flex ASM. You can't shutdown the ADVM Proxy unless ACFS is running.
If the instance crashes, the MGMTDB fails over to the available node.
Recent Posts
How Rackspace Leverages AWS Systems Manager
October 9th, 2024
Windows Server preventing time sync with Rackspace NTP
October 7th, 2024
Building a Google Cloud Landing Zone in a Scalable, Repeatable and Secure Way (Part 1)
August 28th, 2024