I have a two question regarding the following lines in the
error log

01:00000:00075:2010/09/21 03:02:56.00 server The 16K memory
pool of named cache AdminDbData (cache id 1, cachelet id 1)
is configured too small for current demands (state 1).
Transaction progress may cease or response time may

1. What does it means?
On that time ( 03:02:56), a query that needed a room in the
cache could not find place ?
Is it since the checkpoint could not keep up to remove dirty
pages to disk?
If cache work on LRU, and old pages needs to go out, why the
cache did not remove old pages and got the new pages?

2. Then, what?
The specific query that had no room on the cache, needs to
do what? wait? do I/O? is it missed query?
When the message appears,what happened to the query that
triggered this message?