Re: Segment management auto clause

Here's another experiment to do some time.

I think you get just one L2 block per segment until the segment is very large.

So if you have two RAC nodes, and you are using ASSM to avoid collisions (and ignoring the inevitable collisions on index leaf blocks) then a process which insert rows very rapidly will be pushing L1 blocks to FULL very
frequently.

But the L2 block keeps statistics on the numbers of L1 blocks it controls - if these statistics are updated every time an L1
block is filled, then the L2 block could become the one block that fires back and fore across the interconnect - just like the segment header block in OPS before
anyone though of using multiple freelist groups.

--
Regards

Jonathan Lewis
http://www.jlcomp.demon.co.uk

  The educated person is not the person
  who can answer the questions, but the
  person who can question the answers -- T. Schick Jr


One-day tutorials:
http://www.jlcomp.demon.co.uk/tutorial.html

____Denmark__May 21-23rd
____Sweden___June
____Finland__September
____Norway___September

Three-day seminar:
see http://www.jlcomp.demon.co.uk/seminar.html
____UK_(Manchester)_May x 2
____Estonia___June 4th - 6th
____Australia_June 18th - 20th (Perth)
____Australia_June 23rd - 25th (t.b.a)
____USA_(CA, TX)_August

The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html


"Jonathan Lewis" <jonathan_at_jlcomp.demon.co.uk> wrote in message
news:b9ua3g$bjo$1$8302bc10_at_news.demon.co.uk...


>

> You might try running series of single row inserts on tables

> at that size, after doing some scattered deletes.  Check how

> many different blocks of what type get latched - the behaviour

> I've seen on smaller tables is:

>     seg header

>     L2

>     L1

>     Table block.

>

> Which makes for 2 very hot blocks, and a handful of warm (L1)

> blocks when the going gets tough - instead of one hot freelist

> block.  I'd be interested to see how this changes with size

> of table.

>