Comments on: Intelligent Stats http://jonathanlewis.wordpress.com/2007/08/20/intelligent-stats/ Just another Oracle weblog Sun, 19 May 2013 12:40:29 +0000 hourly 1 http://wordpress.com/ By: Can an explain plan on a SELECT self deadlock? Yes it can … but it is a bug « Christian Bilien’s Oracle performance and tuning blog http://jonathanlewis.wordpress.com/2007/08/20/intelligent-stats/#comment-24102 Thu, 08 Nov 2007 11:49:10 +0000 http://jonathanlewis.wordpress.com/2007/08/20/intelligent-stats/#comment-24102 [...] . Jonathan Lewis also picked it up http://jonathanlewis.wordpress.com/2007/08/20/intelligent-stats.There is however in my (reproducible) case no call to dbms_stats. As some of the related bugs refer [...]

]]> By: Boneist http://jonathanlewis.wordpress.com/2007/08/20/intelligent-stats/#comment-19856 Wed, 29 Aug 2007 13:22:13 +0000 http://jonathanlewis.wordpress.com/2007/08/20/intelligent-stats/#comment-19856 Hi Jonathan,

We had problems with the “Cursor: pin S wait on X” today, but under slightly different circumstances to the post you linked to. I thought you’d like to know about it, for completeness!

We’ve recently upgraded to 10.2.0.3 from 10.2.0.2, and we’ve been having some problems with long-running queries (not related to the upgrade) for a while. So, we got the DBA’s involved today, having come to the end of our own investigations.

He tried to trace the session using DBMS_MONITOR.SESSION_TRACE_ENABLE, but couldn’t get any results due to the session already running the query. Anyway, a while later, we noticed that the query was hanging with the aforementioned wait event.

Some digging around on metalink came up with this document suggesting dbms_monitor was the cause, and suggesting that the root cause might be use of SQL_TRACE.

There is a patch for this issue, but we’ve not had time to try it out, yet.

]]>