flashback database: any instrumentation/parallelism?

  • From: "Jeremiah Wilton" <jeremiah@xxxxxxxxxxx>
  • To: "'oracle-l'" <oracle-l@xxxxxxxxxxxxx>
  • Date: Tue, 5 Dec 2006 15:37:21 -0800

In 10.2.0.1 is there any way to monitor the progress of 'flashback
database'?  From what I can see, the session just waits on 'control file
sequential read' counting up to ~40 seconds in wait then starts back at
zero.  There is nothing in v$session_longops.

Meanwhile in strace I see the server process doing lots of work, including
many reads against the flashback logs and datafiles.

Also doesn't this thing work in parallel?  The process can certainly take a
long time.  You can barely utilize the available I/O and CPU on a typical
host with just one server process performing the flashback.

How can it be that there is no real way to see how much is done/remaining or
how it is performing?

Is this just a really immature feature?

Jeremiah Wilton
ORA-600 Consulting
http://www.ora-600.net


--
//www.freelists.org/webpage/oracle-l


Other related posts:

  • » flashback database: any instrumentation/parallelism?