Gotcha. I was scratching my head, I knew I was missing something
Ok, all i have to do is download your github, edit the collectd/collectd.conf.d/network.conf to the IP of my VM and push it to resin.io, right? I haven’t done a docker compose/multi docker setup on resin.io yet, do they ask which yml to take? Or do i need to delete one (the prometheus yml)? Sorry but the process to setup resin.io with a multi docker setup is new to me… thank you very much!
Apologies if there isn’t much documentation yet…
You don’t have to update anything in the repo source, you can set the IP in resin.io – e.g.:
However, you cannot specify the
docker-compose file in resin, so indeed, you need to remove or save the original
docker-compose.yml and rename
If you don’t have anything yet for reporting, you try this on one of your local machines (this is what I currently use)
Thanks. I just did a local install of prom on my Mac as another source to Grafana.
I’ll look at the docker approach when I get sole way through my to do list!
Here’s a shout-out to @Charles. This little guy works very nicely with the no-diode method.
Stick RFM to board, connect to Wemos D1 mini, that’s all!
Love to see this breakout working with nothing except RFM95.
Since some WeMos power are sometimes not excellent quality, I suggest to put C2 10uF capacitor. If you don’t like SMD you can but a classic one (through hole) between 3V3 and GND of I2C connector. Can even be more than 10uF, no problem.
BTW where to you get you SMA antenna 90° connectors, I need them
I have a classic 10uF here, easy to do. What exactly do I gain when doing this? I do not notice many problems only sometimes I need to to a power cycle to restart it. Is that the issue it fixes?
The connector…I cheat…it is RP-SMA
-edit: Somebody slap me, not RP-SMA. An old connector I salvaged from my pi single channel gateway, bought that over a year ago somewhere…
Sometime when you transmit, it can have a power demand that the original WeMos may have problem to give going in a reset (in rare case and on bad board design)
This 10uF one may solve this, but not lockup you mention
The one you mention could exist while polling SPI bus (since you use no diode). May be recall a SPI.begin() before sending to initialize it back is a way to try. One again the lock could also have nothing to do with my fork and SPI it’s just an idea.
Reset is no problem as long as it properly resets I will just catch the next frame.
I am also not very sure if it even locks up but sometimes when I don’t het a ping and reboot it next time I do get one. Can also be a coincidence.
These cheap SMA connectors (I always buy the cheapest ) are made off an alloy which starts rusting after a while outdoors.
So I prefer spending a little bit more now on future projects.
better quality 2 pcb edge
testing readabllity , e paper displays have no backlight - inspiration
1.54 / 2.9 / 4.2
two colour e paper displays have no partial refresh, that means that you have to refresh the whole screen.
but they are the best looking, so if your data doesn’t change to fast they are perfect and use very little in sleepmode.
fastest screen sofar is the 2.9 but I’m waiting for the single colour 1,54
oh… and don’t think in connecting them to a ‘normal’ arduino
No problem running up your fork on the Raspis, I just did a local install of Prometheus to point to them and voila.
These seem to be the most useful metrics - any other you’d recommend?
I have something very similar – this is my playground:
But I spent more time to get the environment right than anything else, so it is still work-in-progress…
I’m about to mount an external colinear on my roof. I’m curious to see what difference it makes to my reception (compared to my dev gateway which is downstairs with a small external ant).
ePaper Display, I like … just ordered one
Actually 3d printing with a Da Vinci mini would be better.
Your Bla Bla seems a little high
I know … I have to println that in RED