Wireless network repeatedly going down - Z1 also dropping, not as frequently

[CWK_Rob]

Hi folks,

I've got a really simple branch deployment in a construction site trailer.  Z1->Unmanaged POE Switch>2 APs.  Flat network - generally only 3-5 clients at a time

See the attached screenshots - my APs repeatedly go down throughout the day (less so if there's no traffic).  When the APs are in a particularly bad state - the Z1 will go down too.

I'm thinking there may be a broadcast storm caused by the unmanaged switch?  Anyone experience this before?

[Ryan_Miles]

Can you try a factory reset on the Z1? Poke the reset button with a paperclip or whatever. Hold in for ~10 seconds then release.

[CWK_Rob]

Yeah, it kind of looks like the firmware cycle.  I had a couple of MX's earlier this year that failed during upgrade and the same thing happened.  I'll try that for sure.

Will that explain the AP outages?  They seem more persistent than the Z1 dropping

[Ryan_Miles]

I checked it again. It's running the highest supported firmware version a Z1 support (14.56). I don't actually see it rebooting with any scary messages related to firmware, panics, etc.

Running the ping tool against 8.8.8.8 I see a fair amount of loss here and there along with some fairly high latency spikes. If you bypass the Z and connect directly to the modem is it stable? Mostly looks like an upstream issue from the Z from what I can tell.

Is the modem ethernet port 1 gig capable? The Z is showing the WAN port being at 100 Mbps at the moment.

[CWK_Rob · Accepted solution]

So, I found an old cable that was being used to uplink the Z from the ISP - replaced it and its now running at 1G

I also replaced the unmanaged switch with a managed (non-Meraki solution)

So far so good.  If I continue to run into errors my next steps are to factory reset the Z1, or connect my switch directly to ISP and bypass the Z to confirm.

I've got a feeling (fingers crossed) that the managed switch will do the trick.  I noticed that before the Z1 drops, the APs are down for quite a while.  The APs went down as soon as staff started to arrive this morning (first clients on network since friday), then the Z1.  I think the unmanaged switch was creating a loop between the two APs.

Time will tell - I will update

[CWK_Rob]

OK, it's been stable since swapping the switch yesterday.  No drops on the APs or the Z1.

Solved!