Why have I lost smart half hourly usage data on Daily usage section of online account?

Erm… yes, thanks for spotting that grammatical error @Tim_OVO  I’ve amended it above!

Did the developers check if any changes had been made to the uSmart software on 17th Dec?

We need to remember that the new Billing System doesn’t directly communicate with DCC. It’s uSmart provides the BOL (Business Orchestration Layer).

Morning everyon!

I’ve been advised that:

“An issue was identified where most of the gas meters stopped sending half hourly (HH) consumption in volume units. This was fixed back in December where the meters where re-configured to start sending again HH consumption in volume units.”

I think we may have found our candidate. Let’s test our current way of passing on feedback to these devs: @SianiAnni @RichH if this HH gas usage data goes again, let us know here. I will grab your account details and go direct to the devs to check if this correlates with any of the changes they’ve made, or if it’s related to uSmart.  

Thanks for the explanation, but there’s insufficient detail for me to feel confident that the issue has been addressed.

Think about this for a moment:

Here’s my gas readings for the month of Jan, which is after that bug was fixed.

So there are obvious (and different) points of “missing data”.

Let’s look at the HH readings for both of those days:

A: For 2nd Jan, the HH totals bear no relationship to the bar on the month-histogram. Since there is no correspondingly high reading the following day, should I assume that this low figure for the day has been applied to correct previous days when the usage was over-reported?

There’s no indication that any “correction factor” has been applied however. This leaves the customer unaware of what’s happening.

The other “issue” is the final reading of the day, which appears to be missing. Here’s the same graph obtained from my Comms Hub by Bright:

Note the 48th reading of the day is present.

That data is available to OVO, but still hasn’t been collected.

B: For 20th Jan, there certainly are HH readings up until 17:30 and yet no day-reading has been shown on the month histogram.

Tim and I had a hypothesis about the reasons for this. But we had to abandon that theory a couple of months ago. It concerned the presence or absence of that final reading of the day.

Once again, here’s that same data from Bright:

Note that 48th reading is present.

Let’s stop there and allow @Tim_OVO to show this to the Orionites.

It takes a long time to collate this level of evidence to be fed back. So I want to ensure it’s being noticed.

This is incredibly useful feedback, @Transparent - the time and effort required is not lost on me, so thank you. 

I’ve passed this on, and sent a more detailed updated vis the Super User Treehouse, here

One of the issues to consider is the time after midnight at which the SMETS_command is received from OVO to retrieve the readings for the previous day.

As the gas meter only awakens from its slumbers to transmit data every half-hour, there is no guarantee that the Comms Hub will have received the 48th reading until 00:30. Retrieving data prior to then risks the final reading for the day being absent, even if the ZigBee link from the gas meter is perfectly functional.

Hi everyone, 

Following on from updates from the likes of @SianiAnni and @RichH that confirmed that missing gas half hourly usage date had been populating in online usage graphs since mid December (and subsequent reports from @Transparent that indicated the opposite), I’m looking to collect more feedback from those that were previously effected. 

I could set up a private area and give willing members access, to allow this feedback gathering to happen free from concerns of who else is watching. So if there’s demand for it, let me know @Jequinlan @Thurion @PRoberts @TomatoPlant @TomThumb @TheBatsWhiskers @Steve77 @Blastoise186 

I think it might be useful to have one, even if it’s done on an open invite basis. It could also allow more in-depth discussions to take place without clogging up the main public areas too much.

Well I think we’re chasing more than one “bug”. We might only be able to discriminate between them by sharing symptoms that are specific to our personal accounts.

So if we’re wanting to capture useful feedback from a number of customers, then I’d favour the use of a Private area (aka Treehouse). Otherwise some might be reluctant to post details on such an open Forum.

The other advantage of a Private Area is that those who agree to be within it should be assured that this acts as a replacement service for faults that would otherwise be notified individually to Customer Services. Whilst we’re bug-hunting it’s very important that an over-enthusiastic staffer in CS doesn’t send a Reset-command to our site, thus destroying all the symptoms we’ve been so painstakingly reporting!

Once we’re in a private area, the OVO/Kaluza techies are basically agreeing that our reported issues will be fixed, but in the meantime we try to leave code as undisturbed as possible during the analysis phase.

I’m willing to be in such a group, if it will help. 

Thanks for confirming @Thurion .

Just so you know, the forum software that we use can be a bit quirky when your permissions are changed, and it often won’t fully apply the changes until you log out and back in, which makes sense.

The snag is that it has been known to instantly kick people out if they were logged in at the time, but you should be able to log back in afterwards and the new permissions would then be active. It’s something that has been forwarded to the provider and hopefully should be fixed eventually. Please don’t panic if this happens!