On 1.6.1_RC2 we have just had a panic. A trace shows the following:

extent_alloc_subregion1
extent_alloc_subregion1
uvm_swap_allow
uvmpd_scan_interactive
uvmpd_scan
uvm_pageout

The computer has 256MB RAM and a 1GB of swap space. It is running a
heavily used squid proxy and has 2 bridged fxp cards.

Now, I'm also running out of NMBCLUSTERS which I'm about to patch. Is
the above panic linked in at all with mclpool exhaustion or do I need to
look elsewhere?

-- 
Stephen