Change the IP address of 8630MCU and observe for half a day. Failure fixed.
The background network: GK＋RMCC＋8630MCU＋VP8032+vp8030, all the terminals are configured in internal network. No matter the background or the terminal sitecall, the frame is freezing in about 30 minutes or 1 hour. There is data being transferred instead of being received in the terminal.
Got the fault call in the morning on 23rd, and arrived at the fault field at 4:30 that afternoon.
1． We doubt some terminal is in fault so that we schedule any two conference sites to build a conference and the failure also exists.
2． We doubt it is the network that has resulted in the failure. Schedule some conference site, and all the terminals can be reachable by Ping without losing package. Connect the two terminals and 8630MCU with the same switch to schedule the conference, the failure occurred in 30 minutes again.
3． Ask for the support of the research engineer and make sure the version matches. There is no such failure in this version. Downgrade 8630MCU to version 3.09.1 with the parameters of GK and RMCC configured and the failure still existed. When upgrade to version 3.17.0, there occurred the same failure.
4． It was found there is code flow of the figure being sent to 8630MCU while there is no code flow back to the terminal with the remote assistance of the department research engineer in 24th. The media network interface can be reachable by Ping while the 8630MCU can’t connect with any terminal. The engineer confirmed that the software version, hardware version and the parameters are all correct. That indicated there may be something wrong with 8630MCU.
5． Replace the 8630MCU with a backup with the backup was sent in 25th, the failure occurred in 3 hours.
6． Pull out the network cable when the frame is freezing and connect it with the PC. It found that any address is unreachable, and there occurred the information of the collision of IP address. We confirmed that the IP address was being using by an audio device which has been broken for some days and came to work recently. So was the failure of the video conference system.