I think its safe to say that getting up and having to be at Moscone West by 9am was pretty difficult after the big party last night. I managed to drag myself out of bed and was only a few minutes late for the session.
Sessions I attended:
Oracle Advanced Compression in Oracle Database 11g
This was a really good session. Daniel Morgan and Kevin Closson were the presenters. Kevin pretty entertaining and did a great job presenting the material. The material for the presentation changed slightly since the abstract was posted due to the release of 11gR2 in September. The new title was now Hybrid Columnar Compression. You can download these presentation from Morgans Library.
The session started with why you may need to compress data, a brief history of it since its inception in 18.104.22.168 and how it works underneath the hood. There are many new features in 11.1 such as compressed tablespaces, segment compression and advanced options such as data guard network compression, data pump compression, improvements to rman compression, oltp table compression and securefile compression and deduplication. 11.2 brought some changes to segment compression allowing you to compress for OLTP, Query or Archive data.
The presentation continues to talk about Hybrid Columnar Compression, which is only available on the Exadata V2, and how it compresses the data. It also has two new features, Archive and Warehouse compression.
Warehouse compression provides you with 10x average storage savings, 10x reduction in scan IO and is optimized for speed.
Archive compression provides you with 15x average storage savings but customers have reported upto 70x on some data. This is primarily meant for cold or historical data and is optimized for space.
Slide 6 of Kevin’s presentation contains a table which shows how many CPU cores are required to fully saturate the IO channels as compared to a generic system. What surprised me was that with Exadata is that 112 cpu cores could be used for data decompression and filtration!
Upgrading to Oracle Fusion Middleware 11g
We are in the process of upgrading from 10g to Fusion 11g so this session was perfect timing. I haven’t reviewed the upgrade process yet so this session provided a good overview of the process. They talked about how they have worked hard to reduce the manual steps required and tried to automate the process as much as possible. The presentation covered the required upgrade version starting points and then they broke down the upgrade into 4 pieces: WebLogic, Forms/Reports, Portal and Identity Management.
If your upgrading, definitely check out this presentation.
Oracle E-Business Suite Install and Cloning Best Practices
Like other apps dbas, i’ve performed my share of cloning so I wasn’t expecting to learn anything new from this presentation. The Rapid Install and Cloning processes were discussed and it included some great slides on what happens during the cloning process when you pass certain parameters to adcfgclone.pl
For example, when running adcfgclone.pl on the database tier it covered the different actions taken behind the scenes the different parameters dbTechStack, dbconfig and database. I kinda new what each one did but now I have a much better understanding.
They also talked about new virtualization features that may be coming down the pipe. Basically utilizing Oracle VM templates to quickly deploy new environments. Looks interesting
After the session I talked to the presenter and asked him some questions about automating cloning. Once I finish my scripts i’ll post about it.
New Oracle Recovery Manager (RMAN) 11g Features in Release 1 and Release 2
Michael Messina from Rolta TUSC is another entertaining presenter. Throughout the session he asked questions and if you had the answer he was looking for (or not in some cases :) he gave you a tshirt… Thats a pretty good way to get audience interaction at Openworld.
Michael covered both the benefits and limitations of rman. The limitations being backups not including password files, failed backups may leave files around and online redo log backups not supported.
11gR1 features discussed included data recovery advisor, rman proactive health check, better catalog management and duplicate database improvements.
11gR2 features discussed included automatic block repair, set newname clause enhancements, tablespace point in time recovery(TSPITR) and duplicate without connecting to the target database.
Before I left I picked up two books at the Oracle bookstore, 11g New Features for Administrators Exam Guide and Oracle Automatic Storage Management: Under-the-Hood & Practical Deployment Guide. Some light reading for the long plane ride home.