By: Kevin Closson
Nice post, Marc. It’s good to see that folks still discover and document the objects that need to be partitioned with RAC. Good find. Also, Oracle Managed Files was introduced in Oracle9i. I recall...
View ArticleBy: Marc Fielding
The irony is that, had BIGFILE tablespaces not existed, my 768G temp tablespace would already have been spread across multiple tempfiles, and I wouldn’t have encountered this performance issue in the...
View ArticleBy: Yasin Baskan
Hi Marc, Did you try this workload with a temporary tablespace group having several bigfile temp tablespaces?
View ArticleBy: Noons
Hi Mark. I’m interested if this same relationship of busy waits would show when using exclusively ASM rather than file systems? IOW: is it common to all disk/file mappings or is it something specific...
View ArticleBy: Log Buffer #298, A Carnival of the Vanities for DBAs | The Pythian Blog
[...] to use BIGFILE temporary tablespaces in an Exadata system? Marc Fielding [...]
View ArticleBy: Jonathan Lewis
Marc, One of the many items on my todo list is to check to see if there are any significant differences (apart from hitting database limits) between using multiple tempfiles with a single tablespace...
View ArticleBy: Yuri
Hello, Creating the database in Exadata I’m always doing the group of bigfile tablespaces, usually 8. This step completely prevents the temporary tablespace concurrency.
View ArticleBy: Marc Fielding
Hi Yasin: Thanks for stopping by! I did briefly consider temporary tablespace groups, but I noticed a blog post by Riyaj Shamsudeen...
View ArticleBy: Marc Fielding
Hi Nuno, The block that experienced buffer busy waits was block 2, the first locally-managed space management bitmap. Since this same bitmap exists in non-ASM locally-managed tablespaces, I would...
View ArticleBy: Marc Fielding
Hi Jonathan, Interesting question. In my particular case I only tested the single-tempfile case and single-tablespace multiple-tempfile case, and multiple tempfiles did resolve the issue for me, which...
View ArticleBy: Marc Fielding
Hi Greg, I looked back through my notes, and unfortunately don’t have any tempfile size information, nor the original AWR reports. So it is possible that the tablespace was indeed autoextending. Marc
View Article
