Intermittent 603 "declined" on Outbound VoIP trunk

This is bad… I’m testing a Vitelity DID (only OUTBOUND for the moment; inbound calls I’ve left setup via PSTN ZAP channels) and so far, twice over the last 3 days and 300 outbound calls, the outbound calls simply STOP connecting.

When I dial out, I immediately see this at the CLI: -- Called vitel-outbound/1NXXNXXXXXX

then there is a 60 second silence, and then a busy signal, and this is the CLI output: -- Called vitel-outbound/1NXXNXXXXXX -- Got SIP response 603 "Declined" back from 64.2.142.18 -- SIP/vitel-outbound-08ec3c50 is busy == Everyone is busy/congested at this time (1:1/0/0) -- Executing Goto("SIP/440-08eb90a0", "s-BUSY|1") in new stack -- Goto (macro-dialout-trunk,s-BUSY,1) -- Executing NoOp("SIP/440-08eb90a0", "Trunk is reporting BUSY") in new stack -- Executing Busy("SIP/440-08eb90a0", "") in new stack == Spawn extension (macro-dialout-trunk, s-BUSY, 2) exited non-zero on 'SIP/440-08eb90a0' in macro 'dialout-trunk' == Spawn extension (macro-dialout-trunk, s-BUSY, 2) exited non-zero on 'SIP/440-08eb90a0'

However, when I try to call another Vitelity “manged” DID, there is NO delay at all, AND the asterisk “operator” immediately tells me “All Circuits Are Busy Now…”. No busy signal there. The CLI shows: -- SIP/vitel-outbound-09417698 is circuit-busy == Everyone is busy/congested at this time (1:0/1/0) -- Executing Goto("SIP/440-09411248", "s-CONGESTION|1") in new stack -- Goto (macro-dialout-trunk,s-CONGESTION,1) -- Executing NoOp("SIP/440-09411248", "Dial failed due to CONGESTION") in new stack -- Executing Macro("SIP/440-09411248", "outisbusy|") in new stack -- Executing Playback("SIP/440-09411248", "all-circuits-busy-now") in new stack -- Playing 'all-circuits-busy-now' (language 'en') -- Executing Playback("SIP/440-09411248", "pls-try-call-later") in new stack -- Playing 'pls-try-call-later' (language 'en') == Spawn extension (macro-outisbusy, s, 2) exited non-zero on 'SIP/440-09411248' in macro 'outisbusy' == Spawn extension (macro-outisbusy, s, 2) exited non-zero on 'SIP/440-09411248'

There is no activity on the box today (Sunday) as I encounter this issue a 2nd time (the 1st time users contacted me complaining they couldn’t make calls (Fri. during business hours), and this time I was just “checking” out the system). This was the CLI output before attempting to place a call: asterisk1*CLI> sip show channels Peer User/ANR Call ID Seq (Tx/Rx) Form Hold Last Message 64.2.142.30 <MY LOGIN> 145b8cb412f 00145/00000 unkn No 1 active SIP channel

I tried “sip reload”, which had no affect. Issuing a “restart when convenient” at the CLI rectified the problem and outbound calls work as expected once again.

This is trixbox 1.2.3 (Asterisk 1.2.12.1), NAT, behind a wrt54g running hyperWRT 2.1b1 + Thibor 15c.

How can I track this down? Any help is greatly appreciated!

edit: just happened a third time (shortly after I made one successful call)
Where/What can I look for?

TIA