Losing internet causes loss of ping

[BlakeRichardson]

I've had a few internet issues at the office this morning and every time the internet has gone offline and the MS units lose connection to the cloud they stop responding to PING's. The switches IP's are all statically set so it makes no sense why this should happen.

Any ideas?

[PhilipDAth]

Are you sure the Internet is going offline - or is it being caused by the MS units?

As a matter of interest, do they still forward traffic at layer 2 between machines in the same VLAN?

I assume this is the layer 3 gateway configured on the MS not responding?

[MilesMeraki]

I'd make sure it's not a problem within your LAN environment first. What's upstream from the MS switches?

Have you recently upgraded the firmware on these MS's, what are you currently running?

[BlakeRichardson]

The MS unit are all on a pretty flat L2 network. @PhilipDAth yes the problem is related to the internet, our gateway had an issue this morning, I pulled the plug to reboot it and our monitoring software (PRTG and Intermapper) both complained that the units were offline even though the Cisco SB series we have were all still online. L2 traffic still seems to flow.

Everything is communicating via VLAN 1 so there is no L3 routing involved. Its a pretty simple network at the moment. 

@MilesMeraki Its not a LAN issue as only affecting the MS switches which includes is MS-425/32 which is our core switch. 

[MerakiDave]

Odd, I wouldn't expect that. 

I just tried it on a home lab network:

Client---AP---MS---MX---Internet

Started continuous ping from the client to the AP and the switch, then pulled the plug on the MX.

Not a single ping dropped.

Let this run for a few minutes, even after the switch LED turned orange to indicate loss of cloud.

[BlakeRichardson]

Hmm, ok just some more information switch IP's are all statically assigned and even if they weren't the gateway isn't running a DHCP server. Switches are on firmware 9.32

I do have to unplug the backup unit and swap back over to our primary hardware this afternoon so if I remember I might make a screen cast not that it will be of much use. 

[PhilipDAth]

So it is the management IP's of the switches you can not ping from another location on the same LAN?

[BlakeRichardson]

@PhilipDAth Network traffic is flowing fine but when I ping the switches management interface I get no response. 

[Uberseehandel]

---

@BlakeRichardson wrote:

Hmm, ok just some more information switch IP's are all statically assigned and even if they weren't the gateway isn't running a DHCP server. Switches are on firmware 9.32

---

On my test network IP addresses of network devices are statically assigned, by the DHCP set up on the MX (fixed IP assignments for the VLAN concerned). Could this cause a problem? Would it be better to hard code a static IP on the device, rather than assign it?

[BlakeRichardson]

The devices IP's are manually set in the Meraki dashboard not obtained by a DHCP server in any shape or form.

[Uberseehandel]

Well re-reading everything, and knowing that the static addresses are locally set, it sounds as if the management VLAN got mangled \t some point or that there is an implicit reliance on a default/native VLAN whose attributes have changed in some manner.

When the taniwha grabs the wahini, I find it much easier to resolve the problem if everything is explicitly declared.