Two similar setups - one with major drops in qACK on WAN

[focalguy]

Okay, so I've had these two boxes running 2.0 nano since RC3. They are now both running 2.0-RELEASE. They both have a T1 connection on the WAN side connected to an MPLS circuit. On the LAN side there is a cisco router that it handling VoIP for the phones at each location and the cisco uses pfSense as it's default gateway to send any traffic out.

Location A has more phones and slightly more data usage so far. Location B is where my "problem" seems to be. After a few days when looking at the status of the Queues, I see the qACK on WAN has several thousand drops. Looking at Location A which has more traffic in general, I see 0 drops for qACK.

Yesterday I deleted the Traffic Shaper and rebuilt it again using the Single-WAN-Multi-LAN wizard. Today I am already seeing over 150 drops in the qACK. I've compared the details for the WAN and WAN->qACK queues on both devices and they are identical.

Any ideas on what I could check?

[focalguy]

Here are some screenshots of the two locations. After rebuilding the shaper yesterday, location B already has almost 5000 drops on the qACK queue.

[gweidenh]

Are those the only two WAN queues?  Are their any higher priority queues than qAck?

[focalguy]

Looking at the Firewall: Traffic Shaper page
The WAN queues are:
qACK
qDefault
qP2P
qVoIP
qOthersHigh
qOthersLow

In location B there are only drops in qACK and qDefault. In location A qDefault and qOthersHigh are the only queues with drops.

They each have the same queues configured the same way on the same provider's T1.