Re: [HACKERS] path toward faster partition pruning

Поиск
Список
Период
Сортировка
От Amit Langote
Тема Re: [HACKERS] path toward faster partition pruning
Дата
Msg-id 520f8a71-286d-e36d-34cf-363fd74366e1@lab.ntt.co.jp
обсуждение исходный текст
Ответ на Re: [HACKERS] path toward faster partition pruning  (Amit Langote <Langote_Amit_f8@lab.ntt.co.jp>)
Ответы Re: [HACKERS] path toward faster partition pruning
Список pgsql-hackers
On 2018/02/13 20:08, Amit Langote wrote:
> On 2018/02/09 21:36, Amit Langote wrote:
>> 0004-Faster-partition-pruning.patch
>>
>> The main patch that adds src/backend/optimizer/util/partprune.c, a module
>> to provide the functionality that will replace the current approach of
>> calling relation_excluded_by_constraints() for each partition.
>>
>> Sorry, but there is still this big TODO here, which I'll try to fix early
>> next week.
>>
>> + * partprune.c
>> + *      Provides functions to prune partitions of a partitioned table by
>> + *      comparing provided set of clauses with the table's partitions'
>> + *      boundaries
>> + *
>> + * TODO: write a longer description of things in this file
> 
> And I tried to fix that to some degree in the attached updated version.

Here is an updated version.

I realized that 0005 (Add only unpruned partitioned child rels to
partitioned_rels) did that only for (Merge)Append.  That is, it didn't
handle ModifyTable.  I fixed that by teaching inheritance_planner() to do
it.  In the process, I found out that we don't need the
PartitionedChildRelInfo node and related code anymore, so the patch ended
up removing more code than adding it.

Thanks,
Amit

Вложения

В списке pgsql-hackers по дате отправления:

Предыдущее
От: "henry@visionlink.org"
Дата:
Сообщение: Re: Cached/global query plans, autopreparation
Следующее
От: Masahiko Sawada
Дата:
Сообщение: Re: autovacuum: change priority of the vacuumed tables