The Oracle Engineered Systems and other appliances are creating a new subclass of DBAs – the Database Machine Administrator (DMA). The DMA will be working with the hardware, storage, and OS of the machines. This might be viewed similarly to a system administrator, but because it is specific for machines that will house databases and machines that are designed for databases, the knowledge of the databases is valuable. These engineered systems are also not typical standard hardware and other administrators will be looking to the DMA to handle everything from the database, OS, network, storage, and server needs.
There are new systems and appliances that will continue to be added to the offerings so this will not be a complete list but instead provide guidance on what a DMA has as responsibilities.
Note
Even though the Engineered Systems come preconfigured, they need DMAs for specific configurations and server administration.
Exadata Database Machine
The Exadata is Oracle’s first database engineered system to market. Since then it has had a few different changes and releases. These enhancements keep improving the hardware and features of the database. The OS and database do come preconfigured. Storage is available and the InfiniBand internal network fabric for connections between the servers inside of the Exadata. There are currently two different machines with storage expansion, and they have several different ways to purchase the machine for the current need and expand for growth.
The three areas that are going to be more specific to a DMA are OS, storage, and the network. OS is going to be Linux and with maintenance and monitoring, being able to work in Linux is a must. Storage configuration is going to be configured by the DMA for use with ASM and the file system. Network is for the connections to the internal servers and configurations for Real Application Clusters (RAC) .
The DMA is going to perform many tasks that point back to the Systems DBA that we discussed in Chapter 1. The database parameters and configuration, backups, and high availability options are going to be supported by the DMA. Some system changes and patching is going to be handled in coordination with Oracle as these releases are specific for Exadata. Full control is given to the DMA, and they have access to all of the resources in the Exadata. Even though the DMA needs to make sure that the changes, configurations, and coordination happens with Oracle, they do not have to work as closely with the other OS, server, storage, and networking teams. These systems are all contained with the DMA having to manage the full server.
High availability options need to be maintained by the DMA. Knowledge of RAC and provisioning new nodes in the cluster are part of the job, as are using the internal network and being able to configure and manage the nodes in the cluster.
As a DBA, the storage management is normally with the ASM instance. As a DMA it expands further into the device configuration and mappings to file systems and the devices to be used for ASM disk groups.
There is a client utility that can be used to assist in managing the Exadata, exaCLI. This will help in managing database configurations and storage. DBMCLI and CellCLI add to the utility commands for the storage cells.
Transformation into a DMA is a natural step for a Systems DBA. There are a few more areas to learn about with the utilities and managing storage and networking, but the hardware might be the place where it is most comfortable for you, especially if you enjoy RAC, ASM, and having control on the server.
The Exadatas are machines that are engineered for database workloads, but they still provide configurable options. These options are used for performance and security. The DMA must understand how to provide a highly available, maximum performance, and secure environment for the enterprise that owns these machines. The machines are going to be bring into the environment and start using, but without the know-how of what can be configured as the additional benefits for the enterprise might not be available. The same type of understanding for RAC and security with other database servers are needed for Exadata. The advantage is that the DMA is able to make those changes (after a proper testing and change control) but they are the ones in control of the machines.
Along with the Exadatas there are other engineered systems for analytics. The applications can be on these servers, and the Oracle Exalytics is the newest to join the group featuring In-Memory software. These machines are all part of the DMA portfolio. Large companies can have several Exadata machines while others have only one or two. Exadatas can also be used to manage private cloud environments, which may mean there will be more servers coming soon as the migrations to cloud happen. Of course in the large enterprise, there is plenty to do with all of the Exadatas, but the DMA can also support the various appliances as part of the job.
Appliances
Database, Big Data, and Private Cloud Appliances are pre-configured servers. These appliances all require more administration on the database side as well as maintaining the server. Each of these appliances has their own purpose and not having to worry about the initial configuration and software installations takes the guessing out of the setup. It also provides ease of deployment in using these appliances.
The DMA can support these configurations and appliances. Here the knowledge of databases and the OS is needed to provide the proper support for these databases. Again, the DMA can very easily come from the system DBA tasks to support any and all of these machines. Big Data appliances will have other software such as Hadoop and possibly integration tools to work with.
Since these systems, DBA skills are still needed should the DBA need to transform into another position. This is a smaller transition because the database skills of backup, recovery, and high availability are required and then add the storage, OS, and server maintenance work to become the DMA to support the engineered systems.
This is not a book to discuss all of the interworkings of the engineer systems and why to choose one over the other but to look where DBA skills can be used and what systems will need the support in the future. Database servers, appliances, and engineer systems are going to increase their offerings and will need DMAs to step up and be administrators for these systems even if they are in the cloud or Big Data offerings.
Zero Data Loss Recovery Appliance is a recently added appliance. It is for real-time redo transport to protect Oracle databases on-premise or in the cloud. It automates backups to tape from the appliance and off-loads the workload of the backups on the production servers. Monitoring of the recovery appliance is going to ensure that the processes are working and as a DMA, testing restores will be needed.
On the storage side there is the ZFS Storage Appliance. This can provide shared file systems to Exadata, engineered systems, or other servers. This is where more storage and OS management comes to play for the DMA.
Cloud Engineered Systems
The Cloud Appliance is a machine that can be used to create a private cloud offering. The appliance is managed by the DMA and can have mixed workloads. Since this is an appliance, it comes preconfigured with virtualization to stand up a private cloud system in the company’s data center. Being a private cloud, it requires someone or a team to manage it in the enterprise. This is more than just the database in the cloud but can also house other applications, and requires understanding of the application services.
The Cloud Engineered System is hardware that is managed by Oracle on-premise. It provides the same Oracle cloud service but on-premise. This means that the data doesn’t need to leave the data center, but all of the management of the cloud, Oracle does for you. DMA wouldn’t be the ones to ask to manage the Cloud Machine. The Cloud DBA would be handling the databases whether the databases are on a public cloud or on-premise in a public cloud like the Cloud Machine.
SuperCluster and Other Servers
The SuperCluster is another engineered system that varies from the Exadata, but has specific hardware properties that provide additional security and performance benefits. Since this is an engineered system, it is managed by a DMA; however, the OS is Solaris is another opportunity to learn and have another OS in the tool belt.
For private cloud offering, the SuperCluster can be utilized. It is optimized for Oracle database and applications, but is definitely a powerful machine to have access to virtualize for self-service cloud offerings. As we look into private cloud offerings, we will also be looking into more automation and how this will affect the DMAs and DBAs.
Other competitive servers are going to have the same purpose of specific database configurations to provide database infrastructure in the data center. These servers are going to be managed by those understanding all of the backbone of the database: server, OS, network, and storage.
Again, understanding that there are disruptors and new innovation, more servers and cluster developments will continue. There will be new hardware and clusters to manage in the future. Especially with systems that are specific for cloud and database, infrastructures are going to be the challenges for the DMA.
Architecture Decisions
Another challenge for the DMA is significant knowledge about all these offerings and continuing to research any new enhancements. Each of these options has technology, capacity, or placement that will fit different business requirements. The DMA can recommend the right machine for the job. The communication must exist with the business owners to understand what they believe are their needs, and then be able to educate them on options, alternatives to what can drive the future growth, and be the game changer that they are looking for.
Figure 4-1 Oracle machine offerings
Figure 4-1 shows the choices we have from Oracle machine offerings . Some of these machines work together such as Exadata, Zero Data Loss Recovery Appliance, ZFS Storage Appliance, and Exalytics Machine. Other machines are options instead of Exadata and they are based on point of entry, workloads, and size of the enterprise.
Design and architecture decisions come from knowledge about the choices and road maps of the systems. The DMA can dive into these areas with the understanding of what needs to be supported and how it will bring value to the business.
Automation
Simplifying the process with preconfigured machines and machines that are engineered for the databases allows the focus to be on other aspects of the environment. There is maintenance and monitoring along with some administrative tasks that will need to be examined for automated processes. Something that is probably executed twice will need to be automated. The jobs for backup, configuration checks, and storage maintenance should be monitored but not day-to-day tasks because of automation.
Another aspect of the automation is understanding that all of these systems should be supporting self-service. The DMA is enabling these services. In providing private cloud and databases access to the DBAs and Cloud, DBAs should be able to focus on their part instead of infrastructure and server configurations. The ability to be the starting point of these services and allow additional automation provisioning and self-service on top of that is a goal and the value that the DMA brings to the environment. Another goal is to assist others in how to provide the services based on features of the underlying infrastructure so that it is not underutilized.
DMA
The DMA is in control of the database machines, from engineered systems to appliances. Providing the administration for these servers means more than just being a systems DBA. There needs to be understanding in critical areas of the systems. They need to support the OS, network, storage, and database.
The main tasks for the DMA:
Server OS : Oracle Linux, Solaris
RAC and Data Guard
Enterprise Cloud Control
Backup Software (also Recovery Appliance)
Exadata Architecture and configurations
Implement Exadata features or other engineered systems
Storage Cells and ASM
Network and InfiniBand
Provide automation on server side
Support Self-service databases and applications
Knowledge of server offerings
Conclusion
These are opportunities for DBAs to take system roles into an administrator for the database machines that support cloud, databases, and other applications. The challenges are available to grow into technologies and be involved as these engineered systems provide the backbone for the database environments. DBA transformation into a DMA is an exciting opportunity for the DBAs that find interest in the main tasks and see the value in supporting Cloud DBAs and automation for these systems.