9 years ago

Oracle Database In-Memory: Running in the Real Application Cluster environment

Oracle Database In-Memory optimizes both analytics and mixed workload OLTP, delivering outstanding performance for transactions while simultaneously supporting real-time analytics, business intelligence, and reports.

Four Cornerstone has always maintained that one of the best reasons why you should work with Oracle is because you can be sure that your IT investments go a long way. Oracle has hardware and software that work well with each other and often has features that complement one another. So you get more out of your Oracle software and hardware without having to invest more.

Take a look at Oracle Database In-Memory. It can run in a Real Application Cluster environment using a shared nothing architecture. That means that your queries are run in parallel to access in-memory objects to execute the queries.

However, you would be able to appreciate how Oracle Database In-Memory in a Real Application Cluster environment works with Oracle Engineered Systems when you look at using DUPLICATE and DUPLICATE ALL subclauses. The DUPLICATE and DUPLICATE ALL sub-clauses work with the In-Memory Compression Units, which together with the Snapshot Metadata Units make up the In-Memory Column Store. These In-Memory Compression Units stores the columnar data.

Why are DUPLICATE and DUPLICATE ALL important? When in a Real Application Cluster environment, all the data you put into your database are populated into all In-Memory column stores you are working with. When one column store or In-Memory Compression Unit becomes unavailable, you can no longer access the data stored in that unit and you would need to access the data from disk, flash or buffer cache. That is unless you have duplicated or copied these in-memory compression units in another in-memory column store.

You can use DUPLICATE to mirror an In-Memory Compression Unit and mirror it in memory. This way, even if the cluster containing the primary copy fails, the system can still read the duplicate or the mirror. You can use DUPLICATE sub-clauses on object or tablespace levels and you can also use it on a subgroup of your tables, such as duplicating only this year’s partition and not touching other partitions even when it is found in a similar table.

You can also use DUPLICATE to roll upgrades and patches using RAC services.

Meanwhile, you can use the DUPLICATE ALL sub-clause to copy each In-Memory Compression Unit that is available in all the In-Memory column stores in a RAC cluster. This is especially useful for high availability when you need to have all the tables in memory for each instance of your database. It is also very useful if you want to improve join performance when using a star schema between a large fact table that is partitioned and a smaller dimension table.

Again, you can only benefit from using DUPLICATE and DUPLICATE ALL subclauses when you are using Oracle Database In-Memory, Oracle RAC and Oracle Engineered Systems. Trying to implement any of these Oracle products? Call Four Cornerstone at 1 (817) 377 1144 and work with a team of Oracle certified experts who can help you make the best out of your Oracle investments.

For more information, you can read our source here.

Photo by Oracle.

Scroll to Top