X-Original-To: pgsql-admin-postgresql.org@localhost.postgresql.org Received: from localhost (av.hub.org [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id 3D17B52C04 for ; Thu, 18 Aug 2005 14:00:43 -0300 (ADT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 55868-09 for ; Thu, 18 Aug 2005 17:00:36 +0000 (GMT) Received: from host54a.simplicato.com (host59a.simplicato.com [207.99.47.59]) by svr1.postgresql.org (Postfix) with ESMTP id 0AABB52B3F for ; Thu, 18 Aug 2005 14:00:33 -0300 (ADT) Received: from host54a.simplicato.com (localhost.simplicato.com [127.0.0.1]) by host54a.simplicato.com (Postfix) with ESMTP id D36436C07A3 for ; Thu, 18 Aug 2005 13:00:28 -0400 (EDT) Received: from [10.0.0.16] (64-60-192-130.cust.telepacific.net [64.60.192.130]) by host54a.simplicato.com (Postfix) with ESMTP id 801756C0604 for ; Thu, 18 Aug 2005 13:00:28 -0400 (EDT) Message-ID: <4304BEAF.6010308@tentacle.net> Date: Thu, 18 Aug 2005 10:00:31 -0700 From: Kaolin Fire User-Agent: Thunderbird 1.0+ (Windows/20050816) MIME-Version: 1.0 To: pgsql-admin@postgresql.org Subject: long idle connections causing problems? Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: ClamAV using ClamSMTP X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.586 required=5 tests=[HOT_NASTY=0.586] X-Spam-Level: X-Archive-Number: 200508/238 X-Sequence-Number: 18557 Just in the past month or so (no code changes, server changes that I know of, etc, for the last six months), I've started getting errors like: org.postgresql.util.PSQLException: Unknown Response Type 8 at org.postgresql.core.QueryExecutor.executeV3(QueryExecutor.java:192) at org.postgresql.core.QueryExecutor.execute(QueryExecutor.java:100) at org.postgresql.core.QueryExecutor.execute(QueryExecutor.java:43) at org.postgresql.jdbc1.AbstractJdbc1Statement.execute(AbstractJdbc1Statement.java:517) at org.postgresql.jdbc2.AbstractJdbc2Statement.execute(AbstractJdbc2Statement.java:50) at org.postgresql.jdbc1.AbstractJdbc1Statement.executeQuery(AbstractJdbc1Statement.java:233) at org.apache.commons.dbcp.DelegatingPreparedStatement.executeQuery(DelegatingPreparedStatement.java:205) The particular response type will change each page load. Doing a process query will show a number of long-running idle postgres processes for that particular website. Other site pools don't seem affected, but this one is heavier load than the others (still only ten thousand hits a day or so). If I kill all the idle processes for the site, one of them will still remain: 24619 ? S 14:21 postgres: xxxxxxx xxxxxxx 127.0.0.1 idle Which I then have to kill -9... which then requires a restart of tomcat to hook back up. possibly relevant version numbers? psql (PostgreSQL) 7.4.7 Apache Tomcat/4.1.31 Apache/2.0.51 Fedora Core 2, kernel 2.6.9 Any thoughts on what configuration I should be looking at, or if there's something I just dearly need to upgrade, ...? Thank you for your time. :) -kaolin fire -http://erif.org/code/detritus/