💾 Archived View for ew.srht.site › en › 2021 › 20210111-environmental-data-1.gmi captured on 2024-08-25 at 00:11:54. Gemini links have been rewritten to link to archived content

View Raw

More Information

⬅️ Previous capture (2024-06-16)

-=-=-=-=-=-=-

2021-01-11

Progress on Project HausBus

#data

#keepitsimple

Projects

My desk at home is hopelessly cluttered with /projects/ in various stages, ranging from contemplating to just about being finished --- its only one more thing ... you all know this, I guess.

So one of my long term projects is "collecting environmental data on premise".

It started out in maybe 2003 and it went a long way and several iterations of technology. State of affairs last summer was this:

This odd collection has served me well for more than 10 years! The WAF (wifes acceptance factor) was actually quite high, and we gained a bit of insight from this collection.

From the description some of you (assuming there are readers at all) will already have slight doubts whether this author is in a sane state of mind. perl? Wuttt? PGPLOT??? Yes well, pgplot itself used to be a Fortran library, if memory serves me well. It has been replaced with giza, which is not a drop in replacement for how I used it.

However, the pgplot based viewer remained working only, because all the neccessary pieces had survived the upgrade of the /viewing/ system to Debian 9 (stretch). It was clear, that on a newly installed system I would fail to piece everything together. Plus there was a number of shortcomings which were asking for more sophisticated tools.

So last summer I set out to make this whole thing more modern[d]. I placed another controller, with new firmware on my desk, together with a Pi running a new collector plus mqtt, influxdb, grafana and the MagicMirror3 software. Plus display, cables, power supply and whatever is needed for a happy controller's life. And the whole thing failed on me. Miserably so! The whole thing turned out to be unstable. It would run for maybe an hour, or for a whole week, and then it would just stall. Each component still blinking, but the communication would fail. And powercycle would /not/ make the system run again! Very strange indeed.

I spent hours and hours on this. No dice.

Then some day I put everything away and declared unconditional surrender.

But a tinkerer's brain will not give up that easily, right?

So over year end season I decided to keep as much as possible of the existing solution and try /the simplest thing, that could possibly work/ to copy the collected data into a newfangled influxdb. Keep it simple!

This /keeping it simple/ required three iterations nonetheless. But I did get out of this with small changes:

New in the odd collection are these:

None of these new toys are self-explanatory to my sort of brain, but heck, I can show new diagrams without resorting to the perl viewer.

Mission accomplished!

And if I look at this, it is already so much easier to use:

Of course there are minor niggles:

tl;dr: You have to start somewhere. And keep it simple!

Bartender? a Talisker 57North, please, with water at the side. Thank you so much!

Cheers,

~ew

[a] amforth home page

[b] 2020-11-04 Leaking Cistern --- Fixed?

[c] PC Engines APU2 single board computer

[d] for some weird definition of modern, perhaps.

Home