I'll have a look at it - I speak some Python. I'll fork the addon and see what I can do
I have been making a Python MQTT request every 30 mins for almost two years. They usually get moved +/- 5 secs or so anyway without even trying. Not sure how to do it with the API (or even why anyone might prefer that to MQTT).
Apart from last few days and a similar incident last year, it seems a very simple & reliable process.
I may be wrong but I believe it will actually be every 30 minutes from the first attempt so it will be 0, 30 and so on after HA has been started.
But I suspect that is not what the users actually NEED at the moment they get new data just after the half hour changes - if you just up the interval to 30 minutes it will, as you say, be on the 30 minute anniversary of startup.
Ahh, I get it now, I just thought would be better to leave like that so it would hit the API at random times but adding a random delay every 30 minutes would work as well!
Oh yes - from my point of view that would be better - but I know people want their data asap ;-)
As things stand we ignore all but the first request every half hour - BUT I would still prefer if those calls were not even made - they are just chewing bandwidth and cpu for exactly zero gain for you and us.
i've got a working proof of concept that will only actually talk to the Glow API between 0 and 5, and 30 and 35 minutes, with a random delay up to 2 minutes thrown in
i'm just working on making the delay async so the sensor update sleep doesn't hang the whole HA instance
I'm new to HA scripting so don't run this yet, I just wanted to share my progress so anyone more experience in HA than I can tell me if I'm barking up the wrong tree
@ackyaz - fabulous
I don't think my approach will work, having the sleep sync causes HA to lock up as it's waiting for the sensor to return something, async sleep seems to work but I'm getting warnings about the update taking over 10s...
You should be seeing half hour data now
Yes ,it's working for me now. Cheers, Clive!
@jackyaz I have also pushed a pull request to the developer to change the interval to 30mins
So it's a case of waiting now sadly
Is there an issue with gas readings through the API today? My reading is currently 0 for today, despite knowing our heating was on this morning!
We don't differentiate - can you raise a ticket at firstname.lastname@example.org - thanks
Sorry I should have clarified, I'm adding a time query to my API call to only return today's usage, and it's that which was returning 0 for today. Since your reply though, it's all just updated. Will keep an eye on it, but it seems the usage for today only updated at around 2pm rather than every 30 minutes.
Ok your data is not automatically updated unless you call the specific api call to trigger it - you should call it once per half hour
I am assuming here you do not have our CAD or IHD?
Hi Clive, Yes I have the CAD. It worked fine up until today. Still something odd going on, as it changed to 43kwh used at 2pm, but now it's changed to 3kwh, when it should just be a sum of total used today, so don't really understand why it's now gone down!
please raise a ticket at email@example.com
Something odd happened last night.
The PC version of the Bright app (https://bright.glowmarkt.com/dashboard) had several instances of claiming the Glowstick was offline and indeed the app on my mobile phone was saying the same. Some of the time the stick itself was displaying one red light; mostly it showed 3 greens.
Unplugging stick & reattaching it brought 3 green lights back but the apps didn't necessarily behave immediately. Messages on screen suggested I ought to reattach stick to my local wifi but that never works on the PC (Why not ? It would be far more convenient to follow the convoluted procedure if it did !) and although most of the parameters were automatically supplied for the mob app, the final (?) of attaching mob to stick's local network (seems bizarre when local devices aren't able to access data directly from stick and stick was obviously attaching to internet OK - see below) just didn't work as described.
Strangely, the old mob running MQTT carried on almost as normal ! Several of the readings normally recorded within 5 secs of the half hour seemed to have been delayed by up to 2 mins but none were missed completely.
All seems fine this morning.
@KS_Eric It sounds as though your stick stayed connected which is the key objective. As Clive said earlier - do raise individual support tickets for any issues.
Bright is not optimised for use on a PC so it won't always behave as expected - we offer no guarantees and one of the reasons we don't have a PC version is because of the Wifi join which requires the use of a mobile device. Future iterations of our CAD without display will use either ethernet or Bluetooth for internet connectivity.
Good to hear all working as expected now.
Just to confirm, it looks like 30 minute updates were disabled again yesterday at 8:30am ish and we're back to overnight (1am) updates for DCC data. Is that right?
Still seems disabled for me, at least.
same issue here, ive not had any updates since 9am ish yesterday
No we have changed nothing - please raise a ticket at firstname.lastname@example.org
It appears to be broken again for me since yesterday, no data coming in to my Home Assistant setup and the Bright app on iOS is not letting me log in.
The problems I reported on this page still recur sporadically. MQTT less affected than the Bright app but not working normally.
I have emailed 'email@example.com' to raise a ticket.
Good morning everyone - to reassure you, things aren't 'broken'. As you know, I believe, we are the only service to offer the more frequent DCC HH data refresh. It looks as though that isn't working as expected at the moment and when the team are at their desks this morning we'll get it sorted out.
And MQTT is designed for people with our devices - if you have our CAD and having MQTT issues and not getting your electricity readings every ten seconds, please let us know.
I'll post here with an update on the data refresh when we have one.
Happy Friday and stay safe in the storms.
@deskyeti - could you please email support? That is unique to you if you can't login, thank you.
Hi all - hope you are keeping safe - the wind is quite terrifying (watching a huge crane swaying a LOT).
Sorry for the hiccough in our more frequent refreshing of the HH data - that started on Wednesday morning - normal service has now resumed.
Have a lovely weekend.