From: yong huang
 Date: Fri, 12 Jan 2001 13:00:40 -0800 (PST)
 Subject: Checkpoint queue

I'm re-reading R. Velpuri et al's Oracle8 Backup and Recovery Handbook
(not the 8i edition). On pages 246-7, it talks about "checkpoint queue". This
is the n-th time I read it and it finally makes sense. [note]

A little more research regarding thie queue. It appears that in 8i, only 8.1.5
has a Checkpoint queue, as seen from v$sgastat (or its fixed table x$ksmss).
Beginning with 8.1.6, v$sgastat or x$ksmss does not have this entry.
[2004/9 note] Checkpoint queue is seen in v$sgastat or x$ksmss in 8.1.5 and both 9i releases, but not in 8.1.6, 8.1.7 and 10g. I don't know why. Nevertheless, checkpoint queue does exist in all versions. A buffers dump at level 1 is enough to show it (alter session set events 'immediate trace name buffers level 1' and look for "ckptq" in trace file). And v$latch(XXX) views have 'checkpoint queue latch'. It just doesn't show up in v$sgastat in some versions.

Yong Huang
yong321 AT yahoo.com

[note]
Except one point: Checkpoint works from the latest dirty buffers toward the
oldest according to the second bullet on p.247. But p.246 says DBWR writes out
the buffers in ascending, not descending, low redo order.