CAB connection timeout after system mute
I found a strange problem :sometimes I press the "system mute "button , then a few minutes later , a CAB(NOT exactly the same one ) disconnect. Even if I press the "system mute"  to cancel system mute ,  the CAB still CANNOT  come back untill I restart the CAB.
Here are  some messages  from "remote log ":
project cab 0x5001 connected
project cab 0x5001 disconnected
project cab 0x5001 connection timeout
project cab 0x5001 connected
project cab 0x5001 disconnected
project cab 0x5001 connection timeout
piond/mute muted :system mute
Also some error code   from cobranet discovery available (Refer to "cobranet Programmer's Reference "):
2228224  BEAT FLOODED
3932160  RXAUDIO DROPOUT
3997696  TXAUDIO DROPOUT
All the Cobranet devices are PEAVEY MediaMatrix products.
Switch Network is based on "hp procurve switch 5304xl  j4850A"  with  "j4878A"  optical fiber module.

Based upon the CobraNet Discovery (Disco) Error Codes that you have outlined, I highly suspect you have a network issue with the CobraNet Network in this system.
Exactly what and how many MediaMatrix CobraNet devices do you have on this network and how are they configured?
Who is the Conductor?
Are you using XDAB? If so, who is the XDAB Master?
Those network switches offer a log to tell you what is going on in the switches. Are there any errors messages?
How is the network set up? You are not oversubscribing an aggregation port, right?
There is no silver bullet to just magically fix the problem. There are a multitude of ways your system could be configured. We need to know exactly how you have yours configured in order to put the error codes in context.
Also, I do trust that you know this, but more for the edification of anyone else who is reading this:
When using CobraNet Discovery (Disco) the errorCode column should be ignored until the errorCount column shows incrementing values. If the values are not incrementing, then whatever was causing the error code has stopped. If the errorCount column remains at zero (0), there are no errors happening so whatever it says in the errorCode column is NOT contributing to the system's current behavior.

Thank  you very much  for your concern !
First of all , this problem and the previous XDAB problem take place in the same project : Two stacks of two NIONs .
These errors in DISCO only  occurs when I press the "system mute "button .
How can we know exactly who is XDAB master , and how can we specify the XDAB master ?
"10...11"  have the higher priority than "1...2"  ,right ?
May I have your email address to send my NPA files , devices list and network topology schematic to you ?
My email address :
zhujy09@163.com

There are 5 halls.
=======================================
DEVICES LIST:
-----------------------------------------------------------------
Hall NO.1:
NION n3  (4pcs)
CAB 16i   (1pcs)
CAB 8i     (1pcs)
CAB 16o  (1pcs)
CAB 8o    (1pcs)
HP Procurve switch 5304xl J4850A WITH 4 modules:
moduleA : mini-GBIC xl J4878A
moduleB : mini-GBIC xl J4878A
moduleC : 10/100-TX xl J4820A
moduleB : NOT USED
-----------------------------------------------------------------
Hall NO.2:
CAB 8i     (1pcs)
CAB 8o    (1pcs)
HP Procurve switch 2524 J4813A WITH 2 Transceiver modules:
Transceiver (PORT 25):Gigabit-LX Transceiver(J4132B)
Transceiver (PORT 26):Gigabit-LX Transceiver(J4132B)
-----------------------------------------------------------------
Hall NO.2,3,4 have the SAME devices.
-----------------------------------------------------------------
Hall NO.5:
CAB 16i   (1pcs)
CAB 8i     (1pcs)
CAB 16o  (1pcs)
CAB 8o    (1pcs)
HP Procurve switch 2524 J4813A WITH 2 Transceiver modules:
Transceiver (PORT 25):Gigabit-LX Transceiver(J4132B)
Transceiver (PORT 26):Gigabit-LX Transceiver(J4132B)
===========================================
NETORK TOPOLOGY:
HALL NO.1 switch PORT A1HALL NO.5 switch PORT 25
HALL NO.1 switch PORT B1HALL NO.5 switch PORT 26 (as BACKUP link)
HALL NO.1 switch PORT A2HALL NO.2 switch PORT 25
HALL NO.1 switch PORT B2HALL NO.2 switch PORT 26(as BACKUP link)
HALL NO.1 switch PORT A3HALL NO.3 switch PORT 25
HALL NO.1 switch PORT B3HALL NO.3 switch PORT 26(as BACKUP link)
HALL NO.1 switch PORT A4HALL NO.4 switch PORT 25
HALL NO.1 switch PORT B4HALL NO.4 switch PORT 26(as BACKUP link)
===========================================
NOW ONLY HALL NO.1 and HALL NO.5 are used, the other halls are NOT used yet.
During last two days' test, I find that :
(1) The loss of CAB link ONLY occurs after "system mute".
(2)ONLY the CAB in Hall NO.5 lost link ,and the CAB in Hall NO.1 work well.
I checked out the switch's alert-log via the switch's"WEB Interface" ,showing that the network runs well.
Is it necessary to  config VLANs to separate TCP/IP and CobraNet ?
Is it necessary to  config spanning tree ?