Data Missing from Garmin Connect (Moving Time)

There was a bug introduced earlier this month by an update to Connect that seems to have been partially fixed, but there are definitely some issues with it still.

I've found that with a Forerunner 410 there's a problem if I pause and then resume during the session, but there are no issues if I don't. For the run where you see a problem, did you pause it at the 3 mile point?

I've opened a thread logging my issue -
https://forums.garmin.com/showthread.php?358979-Still-some-issues-with-tcx-file-processing

My posting there also contains a link to the thread that covered the original problem and fix.

I still haven't got a reply, but perhaps if more people start reporting this they may look into it.

Thanks Brian. No, I didn't pause it and auto pause is off too. Mine is also a Forerunner 410.

I'll have a look at your link.

EDIT: Had a look at your link and used the Export, Delete, Import function and now I get ALL the elevation data but the 'Moving Time' is still at 0 so even though I get the full Pace chart the first 3 miles is still a flat line at 0. See here:

https://connect.garmin.com/modern/activity/1309293673

In your post are you suggesting that if it is paused and resumed that Connect essentially deletes everything that is measured before this resume? I wonder if it was 'accidentally' paused/resumed at this distance?

I'll be doing some speed work later today which I will try and ensure it's not paused/resumed although I'm not sure how to do that if it's being done accidentally!? Locking the bezel?

Thanks Brian. No, I didn't pause it and auto pause is off too. Mine is also a Forerunner 410.

I'll have a look at your link.

EDIT: Had a look at your link and used the Export, Delete, Import function and now I get ALL the elevation data but the 'Moving Time' is still at 0 so even though I get the full Pace chart the first 3 miles is still a flat line at 0. See here:

https://connect.garmin.com/modern/activity/1309293673

In your post are you suggesting that if it is paused and resumed that Connect essentially deletes everything that is measured before this resume? I wonder if it was 'accidentally' paused/resumed at this distance?

I'll be doing some speed work later today which I will try and ensure it's not paused/resumed although I'm not sure how to do that if it's being done accidentally!? Locking the bezel?

Connect doesn't delete the data preceding a pause, that data is still shown in the splits, elapsed time, route on map etc. However it does seem to interpret the pre-pause data as having zero moving time, and as a result the charts are messed up, particularly when viewing 'over time'.

I've looked at your link above, some of it seems somewhat similar to mine, some aspects are different.

Regarding accidental pausing, I think that would be very difficult to do on a 410. Pause/resume involves the stop/start button on the side of the watch, not the bezel, and those buttons require a pretty firm push so I think an accidental pause is very unlikely. Assuming that you didn't pause in the session linked above, perhaps there are multiple triggers that can cause this issue, only one of which is pausing.

When the original issue emerged at the start of this month, every run I did showed the problem. When the fix went in, everything seemed fine, but there seem to be edge cases where problems are still happening, as both of us can testify to.

Right it happened again today and this time I had to pause as I do a bit of speed work on a road but a car came today and when I uploaded the front bit was missing.

So I have gone to my original one that I posted above and looked closely for changes and found something. It is a closure of a Track tag and then it opened again. After a wee bit of manipulation I have fixed mine.

In my .TCX below you will see the Track tag closed and then another opened. I deleted both those lines to give the new .TCX as shown on the right.



I then saved the file and deleted the current one on Connect and uploaded the modified file. The below two images show the old one showing my Moving Time at 0 (first image) and the new one showing everything correct. Please fix this Garmin as I would prefer not to do this myself every time! If you send me the code for Connect I am available for a fee. ;)

Before changes:


After changes:

Nice debugging work BeOneOptica.

I could see the same pattern in my problem file, and tried the same fix.

It certainly improved things, the moving time was correct and all the graphs were accurate representations of the data included in the file. However modifying the file in that way does change the nature of the data it's representing in a way that doesn't quite give a true representation of what you're trying to record, given that you paused the watch for a certain period.

It's not so obvious from your file because you only stopped very briefly to allow a car to pass, but you can still see that your pace fell to effectively zero for a period in the pace graph (hovering should show it to be a default of 1:00:00/km). For heart rate, if you wore a HRM you'd get shown a heart rate of 0 bpm for that period. Another curiosity is that if you hover over the pace graph, the tooltip text shows the time at that point on the graph. Go to the end of the graph, and you'll see that the graph covers a duration longer than the cumulative time, it's actually the cumulative time plus the duration of the pause(s).

So for any session where the pause(s) are of a significant duration and you modify the file in the way you described, you end up with a graph that's somewhat different to what your intent is when pausing the watch.

So even if we had the patience to hand-modify all impacted files in this way indefinitely, it doesn't get us the correct result. Which is another reason why Garmin should really fix this issue, or at the very minimum acknowledge that there is an issue. Over the past few days we've triaged this issue for them, have identified a particular pattern in tcx files that's causing the problem, and have specified the triggering action that causes that pattern on one model of their watches. In return there has not been a single word from Garmin administrators, even though they are active on this board and have replied to issues that have been raised subsequent to this one. This doesn't look like a particularly difficult problem to fix, it seems clear that whoever produced the fix for the initial issue did not consider that in a tcx file a lap element can contain multiple track elements.

Agree with everything you've said Brian. I'm a relative new runner so I don't do anything fancy with my data other than log it. However, i certainly don't want to modify the file every time I pause on a run. I'm a Linux user and already had to dual boot my system with Window$ to allow me to upload my runs and download my workouts since they refuse to provide anything for Linux.

Still nothing from Garmin. I'll email support later to highlight this issue and thread.

I have uploaded the requested sample files under the following thread: https://forums.garmin.com/showthread.php?358979-Still-some-issues-with-tcx-file-processing&goto=newpost. Thanks for the attention!

Hi GFM1011,

I can only supply one file now as I wrote over the last one so cannot access it any more.

This is a 4.5 mile run that was paused around 13 minutes from the end and after being uploaded to Connect it only records the last 13 minutes after the pause. The Moving Time before that is all 0.

he Track tag closure and open is on line 3558 (to help whoever see the section).

Thanks for your attention on this. Makes me happy :)

Any news on this GFM1011? it's all gone rather quiet...