Path: news.f.de.plusline.net!news-fra1.dfn.de!news-stu1.dfn.de!news.germany.com!postnews.google.com!i12g2000cwa.googlegroups.com!not-for-mail From: "joel garry" Newsgroups: comp.databases.oracle.server Subject: Re: conventional path export with direct=y Date: 1 Jan 2007 15:03:50 -0800 Organization: http://groups.google.com Lines: 30 Message-ID: <1167692630.271027.290290@i12g2000cwa.googlegroups.com> References: <1167533261.136651.323180@i12g2000cwa.googlegroups.com> <0DWlh.30899$Gr2.22904@newssvr21.news.prodigy.net> <1167633898.882485.325140@h40g2000cwb.googlegroups.com> <45992967$0$6463$ba620e4c@news.skynet.be> <1167682356.543668.276030@s34g2000cwa.googlegroups.com> NNTP-Posting-Host: 70.181.182.6 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Trace: posting.google.com 1167692635 31330 127.0.0.1 (1 Jan 2007 23:03:55 GMT) X-Complaints-To: groups-abuse@google.com NNTP-Posting-Date: Mon, 1 Jan 2007 23:03:55 +0000 (UTC) In-Reply-To: <1167682356.543668.276030@s34g2000cwa.googlegroups.com> User-Agent: G2/1.0 X-HTTP-UserAgent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.2) Gecko/20040804 Netscape/7.2 (ax),gzip(gfe),gzip(gfe) Complaints-To: groups-abuse@google.com Injection-Info: i12g2000cwa.googlegroups.com; posting-host=70.181.182.6; posting-account=YRNZ5wwAAAAg-yYjMFwy3FyWUbPiwNdq Xref: news.f.de.plusline.net comp.databases.oracle.server:191544 I've noticed another clue. The schema is loaded with several imports (after being freshly created with some objects loaded necessary for the application). The first import is a table import getting its data from a schema export of another schema. Subsequent tables are imported from table level imports created with the query option. Any missed objects are added from a full=y rows=n export (I'm not sure that that even does anything), and views and triggers are dropped and recreated.. As part of the load process most every table has its primary key updated, and the whole schema is analysed (so I don't think this is a delayed-block-cleanout issue, which of course would be really scary in dictionary tables). The tables exported by the misbehaving direct schema export exactly match those of the first import. The live load has successfully completed, and I did a direct mode schema export with no problems from the command line. Tonight, the script will run and tomorrow I'll know if there is still a problem. I did notice one privilege issue - the script runs under root, and root is in the dba group. I suggested changing that 5+ years and two generations of hardware ago [siiiiigggghhhh]. jg -- @home.com is bogus. "Hookers? Tomorrow? I thought we were doing the Guggenheim." - Michael Crawford