I am wrong, the tiles are also affected. Currently there’s nothing new here:
I concur. BlackVue mp4 uploads of 1-2 weeks ago are not visible as “green tracks”, nor available on OSMiD. Some seem “stuck” in the yellow icon “Map Update” mode, but I cant really differentiate recent vs >4 weeks ago.
My recent updates now seems resolved, but I only checked the Mapillary GUI.
1 Like
Ingesting… since Sunday for me.
Ditto, but now it’s less “ingesting, ingesting, ingesting” and more “failed, failed, failed” 
I wish it gave a reason. I’ve had the odd corrupt image (which seems to be the one thing it tells you about), and I now screen for those before uploading, but for everything else it’s just “contact support”. Is an image bad? Have I accidentally uploaded a sequence twice? Did something just fall over and it would import successfully if retried?
You can use the link to contact support, sure, but if I report the 26 sequences that have just failed, I’ll end up waiting a week or two, and then I’ll get 26 responses asking for details. To which there isn’t really any response other than “it says it failed, you’ve got the IDs, you’re the ones who can look up why, not me.” 〈/rant〉
The stacks of punchcards keeps falling on the floor.
All my sequences uploaded since September 21st had been ingesting until they have all failed this night. So, I have quickly uploaded them again. 
Who knows this was the right thing to do? Perhaps I will have to delete some duplicate sequences now?
No, after two weeks, they will tell you that all failed sequences were purged and that it is too late to investigate.
If the status is “failed” it is always safe to re-upload. For sept 21 and sept 22 we had a specific (unfortunate) issue, see Upload data loss on September 21 & 22
1 Like
When uploading from smartphone, pictures are removed after upload. So, there is not option to “re-upload”.
Boris, does mapillary give any method of tracing which sequence failed to upload? Using mapillary_tools I uploaded a group of video files on ~24-Sep and one (so far) is reported as failed but no identifying reference is given. If I follow the “…contact support” link then a cluster ID is contained in the email template but this is no use to me as I assume it is the internal mapillary ID. If mapillary reported the original filename as uploaded in the failure message then it would be possible to re-upload.
Yes, that is unfortunate at the moment. We have a feature on our backlog to make it possible for you to see the Mapillary folder on Android so you could also back it up yourself (for example to Google Photos, or just locally), but unfortunately that is not yet available. Of course we are trying to make sure this type of data loss doesn’t occur again.
1 Like
Hi Rob, I assume you want the upload of 17:44 NZ time, right? Give this a try:
- Find the mapillary_tools upload history (MacOS ~/Library/Application Support/mapillary_tools, Windows C:\Users\AppData\Local\Mapillary, linux ~/.local/share/mapillary_tools/ )
- Open the json file(s) in there and look for “f6914b65b163af537db893bc9b423b05” in the field “md5sum”
- Your file name should be in “import_path” 40 characters before or so
Let us know if this helps!
Alessandro
Finally, ingestion went to the end. Pictures are now at the “Map update” stage.
1 Like
Thanks Alessandro, that enabled me to Id and re-upload, will have to wait and see if it processes this time.
1 Like
I wonder if processing stopped again or the system simply cannot handle the load anymore. No fresh sequences on the map since almost a week: