Oracle, Technical

Chicago Oracle Users Group Meeting

So we just had the spring COUG meeting. Two speakers this time, Charlie Garry from Oracle and Dean Richards from Confio. Oracle volunteered a meeting room; their offices are in the Sears Tower downtown. They also provided food. (Very cool!) Toward the end – after 5 or 6 people took off – I counted 24 attendees in addition to the two speakers. A lot of them seem to be DBA’s – I talked to one guy who was doing a four node RAC cluster on Sun Fire T2000‘s (with Sun’s 8-core UltraSPARC T1 processors). But of course there was a good number of sales guys too… there was one guy in particular who really seemed interested in talking to me only about what he was selling. I have to be honest… sometimes that drives me nuts. I love talking tech stuff. I don’t mind learning about cool products and I don’t mind talking to sales guys. But when I’m ready to buy something, (1) I’m going to do my own looking around at all of the available options and (2) I’ll find you. Don’t worry, I’ve got like 10 copies of your business card. If only Oracle user groups could be like those really cool Linux user groups back in Michigan. Just a bunch of folks who meet up at a local coffee shop, order pizza and geek out about cool things you can do with computers.

So here’s my thoughts on the speakers. These are just my impressions; I’m sure that others would have different ideas… but FWIW:

Charlie Garry

Charlie’s topic was “Oracle Database 11g”. He went a bit long but did manage to get through all of his slides. I was talking to another attendee afterwards who joked that Charlie had somehow managed to get every single industry buzzword into his presentation. I did think he had some good content though.

  • The two main themes that I remember were “Value Innovation” and “Agility”. Yes, definitely buzzwords… but in particular I do think that “agility” is something that should be important to IT organizations today. As Charlie pointed out most orgs spend like 80% of their money maintaining existing systems and only the leftover 20% on change. And change is driven by many different things; innovation, laws, organizational change, or growth to name a few.

  • He also touched on the cost of complexity and how consolidation can help with this… nothing new here; I’ve been hearing this for years now.

  • There was a slide that said that Data Warehousing is the #1 growth driver in the market – something I recently touched on in my post about the Fidelity case study.

  • He mentioned that Oracle doesn’t market SE and SE1 editions of the database – a valid criticism in my opinion. From an informal show of hands it appeared that less than half the attendees had were familiar with them.

  • There was an interesting side comment (and question during an open panel at the end) about some organization that was getting rid of the “Oracle DBA” job description and replacing it with “Data Resource Managers”. Of course something like this could just be a name change and nothing more, in which case it’s pointless. But Charlie suggested that it reflected a paradigm shift away from managing infrastructure and toward managing information. In fact that was a constant theme during his whole talk.

  • One minor point that he got wrong; he talked about “differential” vs “incremental” backups. Said that every other database on the market does differential backups which take longer each day of the week since they’re backing up every change since the last full. This is wrong on every level. (1) First of all he has the terminology completely backwards; it’s differential and cumulative and he had the wrong definition of differential. (2) Secondly Oracle is of course not the only software package of differential backups. And Oracle does both, not just differentials.

  • He picked on DB2 a little bit, describing a meeting he recently attended where they said that “security roles” would be implemented in the next version of DB2. He said that DB2 for mainframes has had these for 20 years and Oracle along with every other database has had them forever. Now I’m not a DB2 expert… and maybe this is just a jab that’s not meant to be taken seriously. But I’d like to point out that SQL Query Results Caching – a new feature in 11g – has existed for years in MySQL. So I guess Oracle is finally catching up. :)

Overall impression: high-level stuff; the title was a little misleading. He did mention a few 11g new features but it was all in the context of providing “value innovation” and “agility”… and light on the technical details. Good content for decision-makers; his analyst background showed in his thoughts about industry trends and the big picture.

Dean Richards

I don’t have nearly as much to say about Dean. Just that he was great. :) His presentation was about troubleshooting performance problems by using Wait Events and was much more technical. But between a slightly late start and some good questions early in the presentation he got behind and wasn’t able to finish all of his slides. He started with a discussion of wait events including a review of the relevant fixed views in 8i, 9i, and 10g databases. He also gave some sample scripts for querying these views. Then he had six case studies prepared to give actual examples of putting this into practice. Time-based tuning isn’t anything new either; YAPP’s been around for like 12 years now. But it’s always good to revisit this stuff in my opinion.

He only made it through two case studies: the first ended up being a hot block and the second came down to excessive tablescans (db file scattered reads). I really appreciated his handling of the tablescans; he was careful to qualify all of his statements and be clear that tablescans are not always evil. In particular if your query needs to read a large part of the data then it might make sense to just scan the whole table. All of his screenshots were Confio’s utilities (which of course they sell) but he did a great job of focusing on universal concepts rather than just talking about what Confio Ignite can do.

Overall impression: not as polished as Charlie’s presentation but solid, solid technical content. It was my favorite one and I’d recommend him!

And Now For Something Completely Different

On a completely unrelated note my friend Aris suggested that we should change our name to Chicago Oracle User Group And Researchers. Because then the acronym would be COUGAR.

About Jeremy

Building and running reliable data platforms that scale and perform. about.me/jeremy_schneider


One thought on “Chicago Oracle Users Group Meeting

  1. wow,
    how did i miss that. i thought coug doesnt even meet anymore since i’ve not seen an email from them in a long time.


    Posted by Fuad Arshad | May 16, 2007, 7:34 am


This is my personal website. The views expressed here are mine alone and may not reflect the views of my employer.

contact: 312-725-9249 or schneider @ ardentperf.com




Enter your email address to receive notifications of new posts by email.

Join 56 other subscribers
%d bloggers like this: