Meraki: There are errors in 'Get Device data' metric

[TampaSteve]

Kicking the tires on Rabbix to see if it will work for my business (600+ servers, on-prem infrastructure and services, Azure Iaas, 700+ Meraki devices, etc) and I'm not finding many answers when searching forums or when Googling.

I'm facing a cyclical issue on my discovered Meraki devices where they will produce stats and I see valid warnings but will suddenly balloon up to 600 warnings with the message "Meraki: There are errors in 'Get Device data' metric". I suspect this has everything to do with Meraki's API limit but I cannot locate the smoking gun via Rabbix logs. Aside from this I was wondering if I could manipulate the polling interval for Meraki devices to lessen the API call volume and maybe strike a balance.

Also, can anyone share an opinion on Meraki API vs SNMP monitoring and which is best in this case?

[cyber]

Rabbix?

If we talk about Zabbix here.. then yes you can modify all polling times in templates and items
Assuming you are using this template https://www.zabbix.com/integrations/meraki , in templates you can find those "script" and "Http agent" type of items, which do the actual polling. Rest of it seems to be dependent on those pollings. Seems to be done by default in 3m...
Your error seems to be related to "Cisco Meraki device by HTTP" template... It seems to be attached to all discovered devices, so I would assume there might be somekind of network issues to poll data .. maybe ... I have absolutely no experience with Meraki devices...
There can be some info if you look into each discovered hosts item list and (maybe) there is a red "!" displayed after an item called "Meraki: Get device data" ... hovering over/clicking it may give you something...

[jennyfoster]

It's possible to adjust the polling interval for Meraki devices in Zabbix to reduce the API call volume. However, it's also important to investigate the root cause of the "Get Device data" metric error to determine if there are any other issues affecting the monitoring. As for Meraki API vs SNMP monitoring, it really depends on your specific monitoring needs and preferences. SNMP is a more traditional monitoring protocol and provides a standard set of metrics, while the Meraki API allows for more customized monitoring and access to additional data.

[TampaSteve]

LOL @ "Rabbix". I don't know why I keep doing that. Thank you both for the quick replies and guidance. I'll digest your responses and see what I can figure out. There is a decent learning curve to this solution which is a bit overwhelming, but I look forward to putting all the pieces together.

[TampaSteve]

Looks like the volume of 'get device data' errors is from the "Cisco Meraki device by HTTP" template. I've poked around and Googled but cannot figure out where I can see more backend details. Due to the ebb and flow of the error count I still assume this is due to API limits but I want to be certain. Is there a log I can review to see what Zabbix is getting back from Meraki when it makes these calls?

[cyber]

server logfile. But I doubt it will write anything useful there without increasing logging level (https://www.zabbix.com/documentation...untime-control ).. But most probably you will see same timeout there also, just accompanied with a lot of debug info, which process connected where and with what parameters etc... Timeout probably does not give you any data besides the fact it times out...
You can go through that script (https://git.zabbix.com/projects/ZBX/...ease%2F6.4#443) and find out what is the actual call and try to replicate it with postman or something...

[TampaSteve]

I spun up Manage Engine's Ops Manager Central to see what it was like and it too faces a similar problem. The devices will remain in a green state for about 15 minutes at which point polling begins to fail and they never come back. Very annoying situation.

Thank you again for all the assistance!