Re: text fields and performance for ETL
От | David G. Johnston |
---|---|
Тема | Re: text fields and performance for ETL |
Дата | |
Msg-id | CAKFQuwbaVh+-gzTCnD654mqbdkGmjDKETV2M3oBtCO-f2mKzsA@mail.gmail.com обсуждение исходный текст |
Ответ на | text fields and performance for ETL (PG Doc comments form <noreply@postgresql.org>) |
Список | pgsql-docs |
Dear David,
> The goal in our docs is to point out that using an arbitrary length specification is not required in PostgreSQL.
Then it should say so. Because unexperienced reader then uses this limited focus and generalizes.
It is for the very same reason that progreammers don't perceive the need to limit the string size to its realistic size that ETL processes
are slowed down.
For example a currency where 3-char encoding is used, the field should be char(3) and not text.
> The documentation assumes that the sizes involved here are reasonable for such behavior
On the contrary. When you say "performance is the same.." then you make a wrong impression it is an unversal case. But, if fact it depends.
When you include such table into some dataflow from server A to server B, this process will get slowed down. I explained why.
It is important to teach people that update the model to use logical values that make sense for such cases.
IF you teach instead "ah it is not important, you may use text", then you actually ignore a part of informatic team that provides the usage of this data.
(I refer to SSIS in this context) that provide a very fast dataflow in case there is a known record size.
В списке pgsql-docs по дате отправления: