There’s an interesting conversation about hash clusters growing on the Oracle Developer Forum at the moment. I’ve made a couple of contributions to it, and think it’s worth reading.
For some other comments I’ve made about hash clusters in the past (or future):
-
- also references a link on OTN where I’ve made a couple of comments
Hash Clusters – a costing anomaly:
-
- also references a link on OTN where I’ve made a couple of comments
-
- An introduction to a relatively new, interesting, but (at the time) buggy feature
-
- A closer look at what goes on under the covers.
Sorted Hash Clusters RIP – I don’t think the technology is going anywhere
-
- A closer look at what goes on under the covers.
Partitioned Hash Clusters 12c finally gets them
-
- A closer look at what goes on under the covers.
- Where I compare a cunning rewrite for manually optimising a query with the optimizer’s internal “dirty trick” with sorted hash clusters.
Oracle Scratchpad 
Hi Jonathan,
I think you don’t remember me. But I was the guy in Madrid, who said “Hi” in the middle of the break.
This is not a technical comment. Is just a doubt. Aren’t you an Oracle ACE Director?
Looks like you’re a “pro” in oracle’s forum. It doesn’t really matter, because there’s so many people know who you really are. But it’s really wierd to me, see you just like a “pro”. Anyway excelent blog, like I said before, and I’m testing :
changing parameter “_pga_max_size”, I’ll comment to you my results as soon as posible.
John Ospino Rivas
Comment by john ospino — October 7, 2009 @ 3:12 pm BST Oct 7,2009 |
John,
I’d like to say I remembered meeting you – but in reality I’ve been travelling so much that I can’t even remember when I was in Madrid ;(
I was an ACE before Oracle merged the Fusion Director program with the Ace program. To me that seemed to add a air of commercialism that hadn’t been there before (a bit like linking the technical support department with the sales department), so I took my name off the list. I think the balance has probably swung back to be closer to the original “technical recognition” aspect that the ACE program was originally intended to be, though.
Comment by Jonathan Lewis — October 9, 2009 @ 3:16 pm BST Oct 9,2009 |
Got it, about ACE director and marketing. Must be like OCM program.
Well, this Sunday I’ve spend a lot of time reading the complete “index rebuild” discussion. And I have another doubt, Why Are you still having discussion with Don?
I think, you’re playing his game. His want to be touch by Guru like you and Mr. Foot, and both of you are give him what his wants.
Aristotle said: “One swallow does not make a spring, nor does one fine day”. If he wants to see a spring when one query go fast, this is the perfect situation to let him go.
It’s bit sad for me, to see you in battle with nobody, and see you spend time replaying low class stuff. The real “battle” is against real problems, there’re so many problems, like “When do we REALLY have to rebuild the indexes”, I think it’s not a real problem if somebody wants to rebuild indexes every night. Other “problems” or situation needs to understand and to solve are concurrency, undo, redo, buffer cache, the amazing shared pool, etc.
Bear in mind, you are Jonathan Lewis, not nobody like me.
Let me tell you something was awesome. In a Julian’s Advance RAC conference, he said good words, trying to explain something about tuning RACs, and he began with: “I’m not Jonathan Lewis, but I can say …”. Those words come from another guru means a lot for me, because I follow this blog, and probably means a lot for you. I think definitely you’re not in Don’s level.
This is just another opinion, from somebody who wants to know how things works, even the differences between “select 1 from dual” and “select * from dual”, and somebody who really appreciate your work.
Comment by john ospino — October 13, 2009 @ 10:14 am BST Oct 13,2009 |