Last day on the course, realisation of leaving the most excellent Hursley campus that reminded me a lot where I went to University at Keele, and that the migration from DB2 V9.7 to V10 on Tuesday. As promised on Thursday here is the late post for the final day of the course. A lot was covered on the last day so a quick round up
DB2 10 for Linux, UNIX and Windows Bootcamp – Day 4 – What have we done – my prospective
Enhancing Analytic’s with DB2
So the new tools that fall under this section were my main interest in this section, but that’s not to belittle the other work that IBM have done in this area with the improved index scanning and query performance.
You can now create JAVA based table functions that could be called to return the content of a flat file into the FROM part of your query. So this means that you could at the most basic level have something like :
SELECT *
FROM TABFUNC.GET_DATA_CSV('Some File Name') AS (Col1 Integer, Col2, Varchar(20))
It would then return you a table that can be queried and used in a sub-query etc
Continuous Data Ingest for Large Environments
The new INGEST utility, now there was a little bit of confusion over the usage according to the instructor, needing AESE edition of DB2, but the material supplied on the course says it is no additional cost! The hands on lab showed it off brilliantly, with the light ETL ability of the tool, I cant wait to start to use it.
DB2 pureScale
So this is touted as being better than HADR as it is nearly impossible to take offline unless major site failure and now to get round that IBM have developed the product offering to be geographically dispersed with examples that they can give on request of getting it too work over 30 to 40 km. This means that your OLTP processing system can have more 9’s added to the end of your 99%.
Workload Management in DB2
IBM have introduced a new way too regulate your users, along with all the old features there is now a way to use your storage groups to help this out. With a storage group is the new concept of a DATA TAG. This DATA TAG you can give your storage groups a relative higher or lower priority. Therefore the SSD storage groups (this quarters data) could have a DATA TAG of 1 and a SATA drive (this years data) could have a DATA TAG of 5 and a “really” old PATA drive (past five years data) to a TAG of 10. This then means you could have two queries, one that touches the most recent data with the DATA TAG of 1 being set up in Workload manager to run a lot quicker, than the other query which compared this month to a month five years ago, would get limited and “retarded” as it has hit the DATA TAG of 10, which is relatively lower than 1.
Data Concurrency & DB2 pureXML Overview
There has not been a lot change here so I have grouped these two topics together as I cant think of anything to mention here, sorry IBM.
Bootcamp Overview and Exam results
This was very worth going thank you to Mark for presenting and proctoring the exams. It gave a great overview of the new features, and hands on experience of features that I do not get to use at work. I am looking forward to moving tomorrow on to V10 and testing out new features like adaptive compression. I also passed the “DB2 V10 fundamentals” exam, so maybe at some point I will get round to doing the DBA exam.