Which "timeSeries"??

[trendkill]

Hi All,

I am trying to create a script to measure packet loss for from WAN links in python. The script is ready & already deployed working as expected with the logic given. In the output there are 0-4 'timeSeries'. I am a bit confused which 'timeSeries' to pick for packet loss & which is accurate. This data will shown in a graph later used for long term analysis. 

API - https://api.meraki.com/api/v0/organizations/{{}}/uplinksLossAndLatency?uplink=wan1&ip=8.8.8.8

Sample output - 

[{"networkId":"",
"serial":"",
"uplink":"wan1",
"ip":"8.8.8.8",
"timeSeries":[{"ts":"2023-01-31T11:38:20Z",
"lossPercent":0.0,
"latencyMs":5.7},
{"ts":"2023-01-31T11:39:20Z",
"lossPercent":0.0,
"latencyMs":5.9},
{"ts":"2023-01-31T11:40:19Z",
"lossPercent":0.0,
"latencyMs":5.8},
{"ts":"2023-01-31T11:41:20Z",
"lossPercent":0.0,
"latencyMs":5.9},
{"ts":"2023-01-31T11:42:20Z",
"lossPercent":0.0,
"latencyMs":6.1}]}

[sungod · Accepted solution]

This call by default returns data for five minutes, so you have five one-minute samples at the time ts in each sample, they are all 'accurate' for their respective ts. If you use the t0, t1, or timespan parameters you can get data for a different period.

If what you are interested in is long term info to generate graphs, look for performance issues etc., I suggest also look at https://developer.cisco.com/meraki/api-v1/#!get-device-loss-and-latency-history which allows you to specify sample resolution and gives extra metrics.