HCC in Theory

Hybrid columnar compression is one of the more powerful capabilities of the Exadata storage platform. I should mention, by the way, that we can also enable it for you on ZFS* and pillar storage, but originally it was designed for Exadata.

Why is it so powerful? Well, in comparison to Oracle’s previous compression options, hybrid columnar compression really does compress. Everything Oracle issued previously didn’t compress at all. It did deduplication.

ACC is real compression and it comes with a choice of algorithms. You can choose which algorithm you wish according to the amount of CPU required, the time it takes to compress and of course the compression ratio you will get at the end of the operation. The compression ratios can indeed be excellent.

Experience shows that virtually all data will be compressed by at least five or six times. I’ve never seen worse than three times. Twenty times or30 times is, in fact, achievable, is realistic to expect on some OLTP type data.

Performance is of course important. The compression workload itself should have no effect whatsoever on the compute nodes because all the compression is done by the cell nodes. Indeed, performance may well often improve because experience shows that it’s far quicker to read a small amount of data from disk and decompress in memory than to read or write the uncompressed data to and from disk.

On the compute node, the optimizer does know what’s going on. The statistics on the objects will tell the optimizer how many rows are going to be retrieved by a single disk read and it will therefore make appropriate use of full table scans and perhaps move away from indexes.

So that’s the theory. Now let’s see what we can actually achieve and how we do it.

Free Online Registration Required

The tutorial session you want to view requires your registering with us.

It’s fast and easy, and totally FREE.

And best of all, once you are registered, you’ll also have access to all the other 100’s of FREE Video Tutorials we offer!