Yeah I am pretty sure the app details are the problem. Option 2 would be perfect in my case to at least track the cpu and memory of Homey. I don’t care about the app resources only when there are issues (troubleshooting).
Deployed v. 0.9.14 to test now:
Here you can select choice 2: with only Homey CPU and metrics.
Thank you I will test.
Update: as expected the app metrics are indeed the problem on resources. Now changed to Homey resources only makes it drop enough to be usable. Thanks a lot, made a donation for your effort!
@balmli, sorry for my late response.
Good to see your effort improving the app so fast however I finally decided to move away from Homey and migrated 95% of my devices and flows to Home Assistant.
As long my Homey is still operational it will handle a few devices.
Can you tell me some more about this item from the changelog in 0.9.13?
Added flows to enable/disable metrics and set write interval
How does this work? I have some sensors that very seldom change values, and in grafana nothing is shown since the default is to log to Influx on change. I need to log on change and at least once per hour. Is this possible to acheive?
In Grafana you can use the fill(previous)
in group by, and the previous value will be used.
Fill prevous only works if there is a value logged within the time range the dashboard uses. So my previous 6 hours dashboard shows no value even when using fill previous if no vale has been logged in the last six hours.
Hey!
So I was looking for a long term solution to storing and analysing all of the sensordata I have collected and will be collecting. Have been looking into MQTT as well as archive insights apps, but this one was really just too easy to work with.
I easily created a free influxdb cloud plan and hooked it up in homey, and things just started to appear in influx. Impressive!
A couple of thoughts after this PoC for myself.
- There is no way to export “OLD” metrics? As I have about 6 months of sensor data that I am also wanting to keep for analysis. Is there any way of fixing that?
- Obviously I want to keep the data over time. Would it be stupid to keep the data in their cloud? Anyone did the maths on cost there? Should I really get a local instance?
- With a local instance, I still want to keep offsite backup, anyone did that with influxdb?
- There is dashboarding functionality in the influxdb cloud offering. Is grafana a better option?
About the app settings, I selected both zone and name as measurement mode, I am not entirely sure which option is the most flexible in the future. As a first setup, do I gain flexibility by choosing both?
Thanks for a great app!
-
Not at this moment.
-
I have everything locally on a Synology NAS and have dockers on it for Influx, Grafana and MQTT. Do not concern about sizing, influx db’s are more like MB’s and not GB’s. Normally you also select the periode you want to save it. Often default for a year, but you can set the retension policy yourself.
-
The Synology NAS makes backups to a separate harddisk or can backup to a cloud service. Data from dockers is included in backups.
-
I only use Grafana
Instead of a Synology NAS, a Raspberry PI with a 64GB microSD card for storage can be a very good start for influx, mqtt and grafana.
You can do a one time manual export of all insights data using Archive Insights. Not sure how to import those into influxdb, but should be possible theoretically. [APP][Pro] Archive Insights - export all your Insights data to a NAS, FTP or web folder
Thanks for the swift answers. Now that I have my data in Influx and Grafana cloud, I am struggling to figure out the tooling. Grafana to create graphs using the Flux query language. It has shown harder to understand for me than I thought. Probably both related to how I set up the influx to use both zone and name? But it also feels like I am missing some key concepts.
Is anyone able to share with me some nice Flux query to get me starting. Specifically I am trying to query a bunch of my temperature and humidity sensor values.
Most guides I find are not using this new Flux QL for v2, but instead the older version, which is why I am not getting things right.
Learning to use InfluxDB (and TICK stack) bit by bit. And also playing with Grafana. Interested to hear if any of you think the following is logical: add a flowcard (THEN) that allows logging text to a (selectable) Influx measurement.
I currently use Papertrails to log to a syslog server. But that is currently separate from my Influx/Grafana setup. Both syslog and measurements are time oriented, but the syslog is of coarse not a measurement. I’ve not yet learned about Telegraf as a means to get syslog data into InfluxDb (e.g. via Syslog dashboard for Grafana | Grafana Labs), but that also seems a detour to create syslog reports in Grafana tables.
So, how would you proceed with this? And would you be interested in having a way to send text (with current timestamp) to an InfluxDB via a flowcard?
RESOLVED: I incorrectly filled in the name of the organization instead of the ID (the settings page explicitly mentions org ID, but as I had not seen/configured such an ID I misinterpreted as being the name of the org). The ID can be found via the InfluxDB 2.0 web UI: select the user icon in the top left menu bar, then select About. the ID is then shown on the right side.
@balmli , I have successfully used InfluxDb with Homey on InfluxDb v1. I am now switching to v2. Have generated a RW token for the bucket, entered all v2 data in the configuration of InfluxDb app in Homey and it says it is connected. The measurements number on the settings page is nicely rising.
But, I don’t get any data in Influx 2.0 web UI. I have been able to transfer v1 data using Migrate time series data from Manually upgrade from InfluxDB 1.x to 2.0 | InfluxDB OSS 2.0 Documentation. And I do see that data (measurements and values). But I don’t see new data coming from Homey (like value average_1 from measurement homey:cpu_load). Do you have a hint on how I can debug this situation?
Ps. My InfluxDb is running Docker image influxdb:2.0.4 on a Synology NAS. I have mapped port 8086 to a free port and set that port in InfluxDb Homey app settings. I currently use HTTP, not HTTPS.
App works great over here sending data to InfluxDB running on Synology NAS (in Docker) and Grafana showing data (also in a Docker). I am using some virtual devices (sensors) to be able te do some calculations to be shown in Grafana.
I have one issue I run into every now and then. This happens after a power outage (mostly on purpose) - the Synology NAS is behind a UPS so it does a proper shutdown. I’m not sure (yet) if it only happens if also the Homey Pro shuts down.
The issue is that the Homey InfluxDB app stops sending data to the Synology NAS. The information in the settings of InfluxDB app tells that there is no proper connection. A reboot of only the InfluxDB app solves the issue.
My assumption is that either the connection is lost or not properly setup/established when both the Homey and the Synology NAS boot together.
Depending on how often I look at the Grafana Dashboard I discover quickly or after a few hours/days there is data missing and I need to restart InfluxDB app.
Questions:
- Does InfluxDB app already retry to establish a previously failed connection?
If not: could this be added? - The missing data in Grafana/InfluxDB is not resend when InfluxDB restarts. Is there a way to send the data logged by Homey since the connection issue started?
- As a workaround I would be happy to be able to restart the InfluxDB app each night. I don’t want to reboot Homey completely. Is there a way to do that?
Hi,
- Yes the app tries to reestablish a failed connection.
- The app will log and keep maximum 2000 entries if the connection to Influxdb is down, and then write all those entries to InfluxDb when the connection is back
- Yes, you can restart the app, by using the “Restart app” flow (under System).
Thnx for the quick reply!
- Any idea how I can find out why in my case the connection was not established for more than a day?
- I have already 60.000 measurements in an hour of 2. Could this value if 2000 be increased or maybe made adjustable in the app settings?
- Oooops. I was looking in the wrong place. Thanks, could have figured that out myself.
When the Homey is restarted the Influxdb app is programmed to wait until other apps are up and running. Otherwise it can connect to the other apps and their capabilities. So restarting both Synology, Docker, and Homey at det same time might lead to some sort of hang situation maybe …
Increasing the 2000 values is not possible. Also, it takes a bit of memory.
Ok, I’ll keep an eye on it
Is it just me, but find the app a bit unstable! Influx after 24 hrs stops sending data to influx, used both docker and app in Qnap. Restarting influx solves the problem.
Its not unstable with me, but it does need a lot of memory and if you have a Homey with lots of apps then it could be that it is swapping with memory.
I have adjusted the cycle to register values every 30 seconds, that’s more than enough for me.