Re: Thousands of parallel connections
От | Gaetano Mendola |
---|---|
Тема | Re: Thousands of parallel connections |
Дата | |
Msg-id | 41214075.1060903@bigfoot.com обсуждение исходный текст |
Ответ на | Re: Thousands of parallel connections (Tom Lane <tgl@sss.pgh.pa.us>) |
Список | pgsql-general |
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Tom Lane wrote: | Gaetano Mendola <mendola@bigfoot.com> writes: | |>I did last week an Ariadne+Postgresql valutation for the company where I work |>and I learned that |>with 250 MB you can open up to 80 concurrent query |>with 500 MB you can open up to 120 concurrent query |>from now on for each 250MB you can have ~40 connections more | | | That does not add up: the graph can't have a negative y-intercept. | There should be a substantial cost to run the postmaster at all, | and then an essentially fixed cost per connection --- assuming | that all the connections are running similar queries, of course. | You're telling us the first 40 connections require zero RAM. I was not speaking about a single process memory consumption I was speaking in general, and indeed I don't know why but seems the first concurrent queries are less expensive, I was able to "confirm" this rule till 2GB I don't know what there is after. BTW the machine is a single processor with HT enabled. Regards Gaetano Mendola -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (MingW32) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFBIUB17UpzwH2SGd4RAiF7AJ9SFrs+sjcHhNyT4BU9svvBHqmrRgCg7A0w es6qvgRJPiu7XzmJ/zup5gU= =6k1Q -----END PGP SIGNATURE-----
В списке pgsql-general по дате отправления: