X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [64.117.224.130]) by svr1.postgresql.org (Postfix) with ESMTP id 5837CD1B8A8 for ; Fri, 5 Sep 2003 02:25:08 +0000 (GMT) Received: from svr1.postgresql.org ([64.117.224.193]) by localhost (neptune.hub.org [64.117.224.130]) (amavisd-new, port 10024) with ESMTP id 76582-10 for ; Thu, 4 Sep 2003 23:25:01 -0300 (ADT) Received: from news.hub.org (unknown [64.117.224.194]) by svr1.postgresql.org (Postfix) with ESMTP id 6D196D1B52C for ; Thu, 4 Sep 2003 23:24:57 -0300 (ADT) Received: from news.hub.org (host-64-117-224-194.altec1.com [64.117.224.194] (may be forged)) by news.hub.org (8.12.9/8.12.9) with ESMTP id h852OvDf052710 for ; Fri, 5 Sep 2003 02:24:57 GMT (envelope-from news@news.hub.org) Received: (from news@localhost) by news.hub.org (8.12.9/8.12.9/Submit) id h852DBoQ050652 for pgsql-performance@postgresql.org; Fri, 5 Sep 2003 02:13:11 GMT From: "Relaxin" X-Newsgroups: comp.databases.postgresql.performance Subject: Re: SELECT's take a long time compared to other DBMS Date: Thu, 4 Sep 2003 19:13:30 -0700 Organization: Hub.Org Networking Services Lines: 75 Message-ID: References: X-Complaints-To: usenet@news.hub.org X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1158 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 To: pgsql-performance@postgresql.org X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200309/85 X-Sequence-Number: 3451 Thank you Christopher. > Change fsync to true (you want your data to survive, right?) and > increase shared buffers to something that represents ~10% of your > system memory, in blocks of 8K. I turned it off just in the hope that things would run faster. > None of this is likely to substantially change the result of that one > query, however, and it seems quite likely that it is because > PostgreSQL is honestly returning the whole result set of ~100K rows at > once, whereas the other DBMSes are probably using cursors to return > only the few rows of the result that you actually looked at. Finally, someone who will actually assume/admit that it is returning the entire result set to the client. Where as other DBMS manage the records at the server. I hope PG could fix/enhance this issue. There are several issues that's stopping our company from going with PG (with paid support, if available), but this seems to big the one at the top of the list. The next one is the handling of BLOBS. PG handles them like no other system I have ever come across. After that is a native Windows port, but we would deal cygwin (for a very little while) if these other issues were handled. Thanks "Christopher Browne" wrote in message news:m3fzjc58ll.fsf@chvatal.cbbrowne.com... > A long time ago, in a galaxy far, far away, "Relaxin" wrote: > >> Have you changed any of the settings yet in postgresql.conf, > >> specifically the shared_buffers setting? > > > > fsync = false > > tcpip_socket = true > > shared_buffers = 128 > > Change fsync to true (you want your data to survive, right?) and > increase shared buffers to something that represents ~10% of your > system memory, in blocks of 8K. > > So, if you have 512MB of RAM, then the total blocks is 65536, and it > would likely be reasonable to increase shared_buffers to 1/10 of that, > or about 6500. > > What is the value of effective_cache_size? That should probably be > increased a whole lot, too. If you are mainly just running the > database on your system, then it would be reasonable to set it to most > of memory, or > (* 1/2 (/ (* 512 1024 1024) 8192)) > 32768. > > None of this is likely to substantially change the result of that one > query, however, and it seems quite likely that it is because > PostgreSQL is honestly returning the whole result set of ~100K rows at > once, whereas the other DBMSes are probably using cursors to return > only the few rows of the result that you actually looked at. > -- > "cbbrowne","@","cbbrowne.com" > http://www3.sympatico.ca/cbbrowne/linuxdistributions.html > Rules of the Evil Overlord #14. "The hero is not entitled to a last > kiss, a last cigarette, or any other form of last request." >