Feature Request: SNMP Polling of Wireless Health

[cpatelbfm]

I know that the Wireless Health feature on the Dashboard is still in beta, but it would be nice to be able to poll the information displayed for Wireless Health via SNMP whether it's through snmp.meraki.com:16100 or polling it locally from the AP itself. I'd like to see this because the alerting capabilities within the cloud dashboard are severely limited and I think some people would like to have this information polled by their own in-house monitoring suites (Nagios, Zabbix, etc.) to better set up alerts.

Thanks!

[PhilipDAth]

I would not be very keen on this one.

There is no existing standard OID that would cover this, which means Meraki would have to create a custom OID.  Not only do you have the grief of having to use custom OID's on the SNMP manager side, it then means Meraki would be locked into supporting that OID.

At the moment Meraki can freely innovate the Wireless Health feature - being locked into support an OID would limit this.

You would be better off writing your own API to SNMP converter if you wanted more state exposed to an existing SNMP platform.

[cpatelbfm]

If that's the case, would it be possible then to expand the alert features within the Dashboard then? It seems kind of asinine (at least to me) that if this feature gets out of beta and there's no alerting capabilities for this, admins are expected to more or less hover over their dashboard with more frequency than I imagine most of us would care to given other obligations on a day-to-day basis. At least some alerting for when the average latency and number/percentage of failed connections pass a certain threshold would be good to have.

[PhilipDAth]

I could be wrong, but I believe Wireless Health is more intended for diagnosing a problem - aka why is something not working - rather than a pro-active alert engine saying that something has stopped working.

For example, these stats are from one of my customers where their WiFi network is working perfectly.  They simple get a lot of people "trying it on".  Right now, 2/3's of their access points are showing as red or yellow.

Imagine if they got alerts each time one of the access points went red or yellow!

[cta102]

I would have to agree with Philip that it's a diagnosis tool rather than a monitoring tool.

On the whole you will be spammed with failures due to mis-configured client devices or roaming clients which never complete authentication.

The amount of filtering you would probably have to apply to get alerts to manageable limits would most likely also limit any benefits.