Re: Rationalizing EXPLAIN VERBOSE output
От | Bruce Momjian |
---|---|
Тема | Re: Rationalizing EXPLAIN VERBOSE output |
Дата | |
Msg-id | 200203110436.g2B4a5U22724@candle.pha.pa.us обсуждение исходный текст |
Ответ на | Re: Rationalizing EXPLAIN VERBOSE output (Tom Lane <tgl@sss.pgh.pa.us>) |
Ответы |
Re: Rationalizing EXPLAIN VERBOSE output
Re: Rationalizing EXPLAIN VERBOSE output |
Список | pgsql-hackers |
Tom Lane wrote: > Bruce Momjian <pgman@candle.pha.pa.us> writes: > > I can't think of a good reason, but making it a select output makes > > EXPLAIN one of the few things you can't get into the server logs, even > > if you want to. At DEBUG5, you get almost everything about a query. > > ... including the query plan dump, no? I don't see the point here. > > One reason in favor of SELECT-like output is that a lot of user > interfaces are not prepared for large NOTICE outputs. (Even psql > isn't really, since it can't paginate NOTICE output.) Pagination is a good point. EXPLAIN is one of the few cases where the output is clearly multi-line. I am concerned that making explain like SELECT means it is on the one piece of debug info you can't get into the server logs. Query dump can already get into the query logs, but not EXPLAIN non-verbose. In fact, as Peter explains it, NOTICE \g goes to the terminal because it is assumed to be an error. Maybe we need to make psql smarter and only send ERROR/WARNING to terminal, and INFO/NOTICE to the log file. With new elog() levels, seems this is needed anyway. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026
В списке pgsql-hackers по дате отправления: