Bit of a random question but I've had it a couple of times now so need to reach out for some advice.

We have 6x SG500-52P switches in a stack via 5G SFPs which have worked great so far.  Into that we have Cisco 7841 PoE VOIP phones which via CDP get put into VLAN30 for voice while our user desktop lan via the phone gets VLAN1, and this works fine.

However, every few months if we haven't rebooted the stack, the management interface dies and takes connectivity with it so anything using LLDP dies and after a few hours, all low level broadcast traffic like DHCP stops as well.  If we power cycle the stack master everything comes back and works fine.   A question around that issue on here seemed to indicate its a known issue which requires occasional reboots.

Anyway, we've had that happen twice this year and a reboot fixed it, but the last time it happened instead of all the phones coming back online after the reboot, probably 40 out of 200 phones failed to connect.  They powered but wouldn't get put into VLAN30 by CDP so couldn't register.  They weren't isolated to one switch either, it was spread across the stack.   If we repatched a phone to another port in the same switch it still didn't work, BUT if we patched it to another switch it worked.   So after that was resolved a few weeks later I tried repatching some of the affected phones back to there original ports/switches but they fail to work still?   Odd.

All I can see in the logs are this:-

2015-08-28 07:56:40 Local7.Warning 172.25.10.1 1 2015-08-28T06:55:29+12:00 172.25.10.1 NT_GREEN - EeeLldpMultiNeighbours - %NT_GREEN-W-EeeLldpMultiNeighbours: Multiple LLDP neighbours on port gi1/1/48 - EEE operational state is FALSE    

Any ideas what would cause this and why as its pretty frustrating and making me nervous for the next time the stack needs a reboot.