全文索引的维护主要有2个部分的工作:
1.同步:
BEGIN
ctx_ddl.sync_index('CNDYBK.CIDX_IMAGE_TITLE' , '50M');
END;
/
2.优化:
BEGIN
ctx_ddl.optimize_index('CNDYBK.CIDX_STORY_TITLE', 'FULL');
END;
/
注意的是,itpub的网友并不建议对全文索引的表,进行analyze.
我查了一下otn的文档,里面有这么一段话:
There is no benefit to having cost-based-optimization of those queries because the rule-based query plan is ALWAYS optimal. The Cost Based Optimizer can only slow them down. As for the Text (Domain) index, in our example the BOOKS_INDEX index. A context index that has been ANALYZED will participate in cost-based optimization. A context index that has not been ANALYZED cannot provide selectivity and cost estimates, so it will do rule-based optimization. Thus, ANALYZING the domain index can change the behavior and performance of the context index. Whether or not it is "required" is up to your application.
So: ANALYZE the domain index if you want to. Do NOT ANALYZE the underlying DR$ tables respective DR# tables.
来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/4096/viewspace-52482/,如需转载,请注明出处,否则将追究法律责任。
转载于:http://blog.itpub.net/4096/viewspace-52482/