Checking in with an update from our team. The current service interruption was caused by an issue with message queuing, which our engineering team is addressing. We’re working to clear up the backlog of data. The good news is you shouldn’t lose any data from this interruption. Thanks for your patience as we sort this out!
Suggest you tell us exactly what “messaging” is as anything under the normal definition of “messaging” does not explain what we are seeing in real time.
Guessing that this is the type of messages and queue Julia is talking about:
If they were increasing capacity for new technology as the vague statements indicate and the message backlog is still there/trying to catch-up, that does not speak well of the entire upgrade they did.
Yeah, I’m a little more empathetic because my wife’s runs a SaaS company, so I’m very familiar with many of the kinds of issues that can cause a break in a very regular data pipeline, plus all the scrambling required to help the pipeline catch up again (regular layered time-dependent aggregations and backfills) for smaller number of customers.
I just got this email:
Together with your words " The good news is you shouldn’t lose any data from this interruption. "
I opened the data from a couple of days ago:
Clearly information still missing.
Why is that customers have to tell that it is still not correct?
Is there no QC in place that would have prevented this ?
unbelievable…
Looks like I’m missing all the aggregated data for two hours, 9AM - 11AM PST, same as you.
But it looks like the rawish half-second power meter data seems to be there, even for the hours in question. Wonder if they started the backfill / cleanup processes based east coast time
Total Usage
Sat 10/1 60.7kwh
Sun 10/2 64.9kwh
Mon 10/3 48.1kwh
Tue 10/4 62.0kwh
Wed 10/5 62.8kwh
But not to worry, all my data is there
What I don’t understand is how Sense can make such a statement “all data is back” while nobody even checked. So someone in the company said it was so, and nobody even bothered to check ?
Even though a lot of graphs with time& date were given in this thread ?!
I would feel betrayed if i would make such a statement and it turns out it is not correct.
Disclamer: Obviously I never participated in US commercial activity.
See what you did - you made me go back and look again. First is initial view on my iPad.
Second is after zooming around a bit. Looks like there is a data corruption / display issue embedded in Oct 3 for me.
And what does your total usage/generation read for the 5 days in the time frame?
SOLAR
Sat 10/1 85.2
Sun 10/2 91.5
Mon 10/3 42.7
Tue 10/4 85.8
Tells the tale, doesn’t it.
Good thing no data is missing
@JuliaAtSense I can say with 100% certainty the usage for Monday 10/3/2022 is incorrect.
I have quite a few critical devices for my aquariums connected to Kasa plugs (HS300’s) that have a constant 24/7 draw with less than 1w variation. Monday shows a decrease in usage. A decrease in usage would mean devices have failed and would have most certainly resulted in deaths, thousands of damages and a flooded house. Fortunately that is not the case therefore I can confirm with certainty the issue is with Sense.
Not to mention the device level power graphs have filled in seemingly accurately though slow to load and must be zoomed in but the usage data graphs and history do not match up.
Could you look into weather Sense is aware if this and weather Sense sill be taking additional steps to correct the usage data and what actions are being taken to prevent something like this from occurring in the future? When issues persist for a week that effect 100% of your users it may be time for a formal executive statement.
Hi all, our Engineering team has been made aware of this new issue and is working on a solution.
So something that started at the same time, has been documented by the customers from the beginning is a completely different and new issue?
It is not that what ever was fixed was a subset of the original problem?
As a smart person once told me, don’t ask a question you already know the answer to.
Hey everyone, an update for you regarding the data you may have experienced loosing. Our Engineering Team has deployed a batch fix to the monitors for the missing data gaps. Expect those to fill in over the next few days as it works its way through all the monitors.