Hoping anyone can shed some help with this…
We have an fs.com S5850 switch connected on vlan 35, subnet 10.20.35.0/24, to our Meraki stack. Port set to access mode, RSTP: enabled (forwarding). STP Loop guard.
This fs switch has several devices connected, including:
- a linux server. Has no connectivity issues.
- several HyperDeck-Studio-4K-Pro-2 - loses connectivity (cannot be pinged anymore) from the Meraki network (but can be reached by the linux server on the fs switch).
The difference between these is most likely that the linux server regularly has network traffic to the Meraki, while the HyperDeck-Studio doesn't.
Initially when the fs switch is plugged into the Meraki stack, all devices are pingable. After some time the HyperDeck-Studio cannot be pinged anymore.
A port capture on the Meraki port connected to the fs switch shows this:
1 0.000000 CiscoMer_4f:00:02 Broadcast ARP 60 Who has 10.20.35.112? Tell 10.20.35.1
2 0.007416 Blackmag_18:be:9a CiscoMer_4f:00:02 ARP 64 10.20.35.112 is at 7c:2e:0d:18:be:9a
So it seems the gateway on the Meraki stack queries and receives the mac address for HyperDeck-Studio.
We do have other issues of intermittency in the network, too. They generally involve connectivity to devices on connected non-Meraki switches, or VM hosts, eventually losing connectivity.
We have Unique client identifier (BETA — Recommended for your network) client tracking selected (as recommended by Meraki support), but I am not convinced this is correct because we do not have downstream L3 switches.
Our network roughly consists of the following:
- a 3 switch stack MS01 (MS01/1 MS01/2 MS01/3) running L3 routing for various vlans. Firmware: MS 15.18. RSTP root is on this stack.
- 2 connected Meraki switches MS02 + MS03 (no L3 routing)
- an Arista 10g switch connected via two ports, on different vlans.
- the fs.com switch connected to MS01/1 port 48, vlan 35, access mode.