Exadata System Software Release 126.96.36.199.0 is now generally available.
This release follows Exadata System Software Release 19.3, bringing further innovations for smarter infrastructure, improved performance, and smarter management.
There have been over 20 unique software features released in the last year including improved performance for all database workloads. Not only across traditional workloads such as transaction processing, analytics, consolidation, but including support for modern multi-mode applications such as IoT, Machine Learning, gaming, and eCommerce. We continue our tradition of protecting your investment in Exadata, with all new software features compatible across RoCE and InfiniBand systems, and all supported Exadata hardware generations.
In this post, we outline some key enhancements in Exadata software focused on three areas.
If you're the TL;DR type, you can just head to the New Features for Oracle Exadata System Software Release 20.x section in the Oracle Exadata System Overview Guide, or view the webcast by Kothanda Umamageswaran, Senior Vice President of Exadata Development.
Still with me, great! Let's look at some features.
Exadata Secure RDMA Fabric Isolation enables strict network isolation for Oracle Real Application Clusters (Oracle RAC) in multi-tenant scenarios. This new feature, available on Exadata RoCE virtualized environments, allows each tenant residing in dedicated virtual machine (VM) clusters to use RoCE VLANs and security tag enforcement at the KVM host level to ensure database servers in separate clusters cannot communicate with each other over a shared storage network. This feature aligns with InfiniBand partitioning for network isolation in previous generations.
See Using Exadata Secure RDMA Fabric Isolation in the documentation for more details.
To increase log write throughput in high performance database workloads, redo log writes to hard drives are now automatically and transparently stored using Exadata Smart Flash Cache in Write-Back mode on High Capacity Exadata Storage Servers. Freeing up hard disk drive resources for I/O intensive activities such as GoldenGate log mining, log archiving, and RMAN backup and restore. Depending on system workload, up to 2.5x improvement in log write throughput can be achieved.
This feature works in conjunction with the Persistent Memory Commit Accelerator (available on X8M systems), which provides significantly lower latency for the initial redo log write from the database, and Smart Flash Logging, which prevents the occasional log write latency outliers. Smart Flash Log Write-Back boosts overall log write throughput.
Oracle Data Guard primary and standby databases both take advantage of this boost in throughput for online redo log files and standby redo log files respectively. When space resource management is configured for the Exadata Smart Flash Cache, caching of redo log files is included in the space accounting for each container database (CDB) or pluggable database (PDB).
This feature provides a significant performance improvement when creating columnar cache with In-Memory database formats by reading data from hybrid columnar formats that are already in Exadata Flash Cache. This is especially beneficial when multiple concurrent workloads are sharing the hard disk I/O bandwidth. For example, if both a backup that utilizes the hard disk bandwidth and the In-Memory columnar cache creation are running concurrently, Fast In-Memory Columnar Cache Creation will use flash cache, allowing both the backup and In-Memory columnar cache creation to run faster.
See In-Memory Columnar Format Support in the Oracle Exadata System Software User's Guide for more information about the in-memory columnar cache format.
For Exadata X8M RoCE systems, this new feature builds on the existing Exadata Smart Flash Cache feature that preserves data during cell-to-cell rebalancing operations, enabling the Persistent Memory Data Accelerator to rebalance the persistent memory (PMEM) cache automatically to maintain consistent application performance. When a rebalance operation moves data to a different storage server, for example after a hard disk suffers a real or predictive failure, any data that might be cached in the Persistent Memory Data Accelerator in the source storage server will be replicated to the target storage server automatically.
To ensure mission-critical databases are given priority to valuable persistent memory resources in consolidated environments, you can now enable/disable Persistent Memory Data and Commit Accelerators at a database level using two new IORM interdatabase plan directives. Plan directives pmemcache and pmemlog can be set to “off” to prevent the specified database from using the PMEM cache and log, which is otherwise available to all databases. ie.
alter iormplan dbplan=((name=testdb, pmemlog=off, pmemcache=off));
See Using IORM to Control Database Access to Flash and PMEM Resources and Controlling Access to PMEM Cache and PMEM Log in the Oracle Exadata System Software User's Guide for more information.
Other enhancements to management in Exadata System Software 20.1 include:
More information on Exadata System Software Release 20.1 can be found in the New Features for Oracle Exadata System Software Release 20.x section in the Oracle Exadata System Overview Guide.
You can also view the webcast of the Exadata Software Release 20.1 by Senior Vice President of Exadata Development, Kothanda Umamageswaran.
Exadata System Software Release 20.1 is the best upgrade target for customers who want to move to Oracle Linux 7. New Exadata X8 and X8M systems will ship with Exadata System Software Release 20.1. Customers upgrading from previous software releases can head to My Oracle Support Note 888828.1 for guidance on the minimum set of Grid Infrastructure and Database Software versions and patches needed to update to Exadata System Software Release 20.1. (Or you could just talk to Oracle Platinum Services to take care of it.)
A big thanks to our Exadata Development team who continue to provide these new releases to enhance the Exadata platform.
We are always interested in your feedback. You are welcome to engage with us via comments here.