@pkinnison on our MXs we get an alert when it is using WAN 2 as it's uplink due to it thinking WAN 1 is down.  It is a standard alert in the 15.16 firmware.   @trendkill I also think the primary uplink alert only applies to the active device, I did have WAN 1 fail on the failover device of an HA pair and didn't get an alert... ... View more

@githo wrote: @cmr It's the most basic and cheapest switch (tried both netgear and D-link). It's a temporary solution due to cable shortage. When we put these switches on an untagged (access) port (VLAN 40) from our Zyxel switch everything works fine. On the meraki the clients get IP's as the should and they can access the servers. They just can't reach (or get a reply from) our gateway (ICMP), so no internet.     Are the MAC addresses of the clients who are connected to the dumb switch in the arp table of the USG? ... View more

Remove and reseat the internal card, my daughter's Chromebook needs this every few months. ... View more

@jmgwozhere I'd physically separate the industrial part from the commerrcial.  Does the industrial even need internet / remote access?   For the commercial / visitor VLANs, ideally you'd have a L3 switch (such as an MS210) but if you've only got what you have then I'd stack the remaining switches and us the wireless NAT for the guests ... View more

Can't you just block outbound traffic to the Meraki DC subnets on the remote MXs?  I think the MX itself would still be able to talk ... View more

So, do you want to tunnel all but not tunnel other Meraki's that aren't yours? ... View more

If you have an SD-WAN setup with an mx concentrator at the DC with an edge device at the remote site, also set up as a hub, then only the networks advertised from the concentrator will route over the vpn.  Internet access will be local.   Or am I missing the question? ... View more

@githo is the dumb switch truly dumb?  It may have a VLAN1 setting that is causing a mismatch? ... View more

Anyone from Meraki care to reply to this, have the known issues from 25.13 that are not in the bug fixes section of 25.14 been resolved or are they still open?   Thank you 🙂 ... View more

@Bossnine  well for us it is the Cisco 8821 fixes ... View more

@MarcP  we are on 25.14 for 7 of our 9 sites and haven't yet seen the issue in 3 days, so that might be a possibility ... View more

We had the same 26.x issue with iPad (2017) devices where about once a week they would simply stop talking and you needed to reboot the AP to fix it.  Rolling back to 25.13 fixed it. ... View more

Can you try 15.16 on your test MX84 with a few more clients?  We use 15.14, 15.15 and 15.16 as our production firmwares but we only use the enterprise feature set. ... View more

The easiest way to get up/down statistics is from the switch port that the MX is connected to, if you have an MS simply click on the port and you will see traffic sent/received under the status heading half way down the page. ... View more

Out of interest, are you using the 26.x firmware, we had been using it but about once a week we had a whole heap of clients stop talking for a few minutes and then some not coming back until AP reboots.  Downgrading to 25.x fixed it. ... View more

In the lodges we did use an MR30H equivalent with SFP, but also with an analogue port that was presented as a SIP line to the network. ... View more

... we used MR33 equivalents. ... View more

We used GPON for lodges and chalets but I'd never thought of using it in the hotel, good idea!  However, we did use one AP per 4 rooms so... ... View more

MR30C designed for corridors with a long thin coverage profile, cheap as possible as we'd be using them for back of house.  At the moment we have to use MR33s and they only get 1 client going past every so often... ... View more

@NFL0NRfirmware stays in beta until I think ~10% of users install it, with 15.x incrementing rapidly this will take some time! We run 15.x on our production SDWAN but even with fairly regular updating we have some devices on 15.14, some on 15.15 and some on 15.16 so we aren't contributing much to the % of any of them... ... View more

on one of our UK sites we do have some in the 100s but as you noted, most in the lower ranges:   ... View more

Site1 looks to be overloaded.  As you have some significant packet loss this will stop you getting near the full throughput.  Do you have up/down stats for site 1 ... View more

If just one device a 10 port 10Gbe switch for iSCSI (no PoE needed but could be provided for other uses) with 8x copper and 2x SFP+ and decent size buffers to cope with the odd 1GbE connected host.  Of course it needs 2 power connections, would be called an MS210-10Ti and would stack with the current MS210s and MS225s.  Ideally it would be a half width 1U chassis and the brackets would allow one per U or two side by side in 1U. ... View more