Horton

Re: POC, WIP: OR-clause support for indexes

Andrei Lepikhov <a.lepikhov@postgrespro.ru>

View thread
From: Andrei Lepikhov <a.lepikhov@postgrespro.ru>
To: Alexander Korotkov <aekorotkov@gmail.com>
Cc: jian he <jian.universality@gmail.com>, Alena Rybakina <a.rybakina@postgrespro.ru>, Robert Haas <robertmhaas@gmail.com>, pgsql-hackers@postgresql.org, Peter Geoghegan <pg@bowt.ie>, "Finnerty, Jim" <jfinnert@amazon.com>, Marcos Pegoraro <marcos@f10.com.br>, teodor@sigaev.ru, Tomas Vondra <tomas.vondra@enterprisedb.com>, Peter Eisentraut <peter@eisentraut.org>, Ranier Vilela <ranier.vf@gmail.com>
Date: 2024-04-01T06:37:58Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Make group_similar_or_args() reorder clause list as little as possible

  2. Allow usage of match_orclause_to_indexcol() for joins

  3. Skip not SOAP-supported indexes while transforming an OR clause into SAOP

  4. Remove the wrong assertion from match_orclause_to_indexcol()

  5. Teach bitmap path generation about transforming OR-clauses to SAOP's

  6. Transform OR-clauses to SAOP's during index matching

  7. Fix the value of or_to_any_transform_limit in postgresql.conf.sample

  8. Transform OR clauses to ANY expression

  9. MergeAttributes code deduplication

  10. SEARCH and CYCLE clauses

  11. Improve estimation of OR clauses using extended statistics.

  12. Teach btree to handle ScalarArrayOpExpr quals natively.

  13. Revise collation derivation method and expression-tree representation.

  14. Instead of trying to force WHERE clauses into CNF or DNF normal form,

Attachments

On 28/3/2024 16:54, Alexander Korotkov wrote:
> The current patch has a boolean guc enable_or_transformation.
> However, when we have just a few ORs to be transformated, then we
> should get less performance gain from the transformation and higher
> chances to lose a good bitmap scan plan from that.  When there is a
> huge list of ORs to be transformed, then the performance gain is
> greater and it is less likely we could lose a good bitmap scan plan.
> 
> What do you think about introducing a GUC threshold value: the minimum
> size of list to do OR-to-ANY transformation?
> min_list_or_transformation or something.
I labelled it or_transformation_limit (see in attachment). Feel free to 
rename it.
It's important to note that the limiting GUC doesn't operate 
symmetrically for forward, OR -> SAOP, and backward SAOP -> OR 
operations. In the forward case, it functions as you've proposed. 
However, in the backward case, we only check whether the feature is 
enabled or not. This is due to our existing limitation, 
MAX_SAOP_ARRAY_SIZE, and the fact that we can't match the length of the 
original OR list with the sizes of the resulting SAOPs. For instance, a 
lengthy OR list with 100 elements can be transformed into 3 SAOPs, each 
with a size of around 30 elements.
One aspect that requires attention is the potential inefficiency of our 
OR -> ANY transformation when we have a number of elements less than 
MAX_SAOP_ARRAY_SIZE. This is because we perform a reverse transformation 
ANY -> OR at the stage of generating bitmap scans. If the BitmapScan 
path dominates, we may have done unnecessary work. Is this an occurrence 
that we should address?
But the concern above may just be a point of improvement later: We can 
add one more strategy to the optimizer: testing each array element as an 
OR clause; we can also provide a BitmapOr path, where SAOP is covered 
with a minimal number of partial indexes (likewise, previous version).

-- 
regards,
Andrei Lepikhov
Postgres Professional