We have a number of Netgear Managed switches on our LAN all working happily in unison but a number of them are suffering fan failure which apparently is not an unknown problem so we have tried to add 4 new switches to the network and progressively migrate the LAN connections from the old switches over to the new switches. When 2 new switches were added everything was fine, but within seconds of adding either of the other 2 new switches to the LAN, the entire network freezes and hangs all the PC's on it. Internet access dies too. As soon as we unplug the uplink to either of the 2 "problem" new switches everything bounces back to working again. We have reset to factory defaults. All 4 switches are set up identically except their WebUI IP Address, we have even tried setting the IP address of one of the ones we have removed to the new problem ones but still the entire network hangs. Would it be that 2 of the switches are just dud or is there something else going on and how do I find out what it is given that it freezes every computer. Most PC's recover when the last 2 switches are unplugged however some workstations hang so bad they need a forced power off / on to reset them. If you connect the 2 problem switches direct to a PC you can get into the WebUI and play around with it no problems. Any ideas? 😞
The switches do not show anything out the of ordinary happening when I add in an extra switch. All I see on the existing switches is the entire network grinds to a halt so bad that some machines need a reboot as they hang to the point of requiring a forced power off and on. On it's own I can connect to the WebUI when I connect one of the switches on it's own to a PC it seems to behave fine. :mad:
Well I finally figured out what the problem was and have stopped the network freezing and hanging all the workstations for anyone else that finds themselves in that position when trying to add a managed switch. It turns out the the Netgear switches I have cannot handle having multiple machines with the same bridge priority value and let the supposed smart switches figure out which one wants to be the root, instead I had to offset each switch so no two switches had the same bridge priority value then they were quite happy to all interconnect. Lesson: do not reply on the Spanning Tree Protocol to sort itself out across multiple switches, set everything manually. It also made a difference to set BPDU to filtering rather than flooding and each port authentication to multi session. With those settings the network is kicking butt!
