I am running emoncms V 8.3.6 along with emonhub on a raspberry pi with a hard drive. I am monitoring a few circuits and temps in my house. I also run a website, so I have access to shared linux hosting. Recently I decided to start logging my data remotely in addition to locally on the hard drive with the raspberry pi to provide redundancy and access without going through the firewall. I setup emonhub to post remotely and it is working great. I figured I would start out with a longer post intervals on the remote server so I didn't tax my own network, and the website I host. I am posting to the remote server with an interval of 60s configured by emonhub, and I am posting locally every 10s. When I setup the feeds, I set the feed intervals to match this locally (10s) and remotely (60s) for power, and both then are also set to log Power to kWh/d. However, after collecting data for a few days, I've noticed that the kwh/d calculations differ by as much as 25%. For example, on my heatpump today, I measure 12.4 kWh locally and 8.4 kWh remotely. Is this discrepancy due to the 60s interval? The data comes from my raspberry pi so I'm guessing that the timezone difference between servers shouldn't be a factor, could it be? When I inspect the power data it looks nearly identical, which is now what I expected with the intervals being different - anyone know what this is? For example, there are the same number of data points reported by emoncms for both servers despite the difference in interval time. See below:
Discrepancy in kwh/d on two servers
Submitted by sbeausol on Tue, 16/12/2014 - 14:05I am running emoncms V 8.3.6 along with emonhub on a raspberry pi with a hard drive. I am monitoring a few circuits and temps in my house. I also run a website, so I have access to shared linux hosting. Recently I decided to start logging my data remotely in addition to locally on the hard drive with the raspberry pi to provide redundancy and access without going through the firewall. I setup emonhub to post remotely and it is working great. I figured I would start out with a longer post intervals on the remote server so I didn't tax my own network, and the website I host. I am posting to the remote server with an interval of 60s configured by emonhub, and I am posting locally every 10s. When I setup the feeds, I set the feed intervals to match this locally (10s) and remotely (60s) for power, and both then are also set to log Power to kWh/d. However, after collecting data for a few days, I've noticed that the kwh/d calculations differ by as much as 25%. For example, on my heatpump today, I measure 12.4 kWh locally and 8.4 kWh remotely. Is this discrepancy due to the 60s interval? The data comes from my raspberry pi so I'm guessing that the timezone difference between servers shouldn't be a factor, could it be? When I inspect the power data it looks nearly identical, which is now what I expected with the intervals being different - anyone know what this is? For example, there are the same number of data points reported by emoncms for both servers despite the difference in interval time. See below:
remote:
Mean: 1008.5
Min: 11.6
Max: 1844.3
Standard deviation: 661.9
Datapoints in view: 720
Local server:
Mean: 1009.4
Min: 11.7
Max: 1843.4
Standard deviation: 663.8
Datapoints in view: 720
Input is greatly appreciated!