Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[mark_h]

I noticed this morning that many of my relational data fields had lost their data.  This has happened many times since relational fields were added.  Fortunately, it only happened a few days ago so I was able to find the data fields relatively quickly in the backups and restore the field data.

But, as noted above, I've seen this happen many many times, and it's always a major frustration to have to go digging through archives for the data files.

I have no idea what is triggering this data loss, but am wondering whether the devs might add some form a debug warning to try and spot when MC is about to mass delete relational data, which might then help get to the root of the problem? 

Or maybe an optional pop up dialogue "Are you sure..." when a mass change is about to be triggered, so that even if the user wasn't aware it was about to happen they could at least be able to "abort" whatever is about to cause it.

Or something.

[JimH]

It could happen if you kill MC when it's saving data.

Antivirus is also a possibility.

[mark_h]

No antivirus in use.

However, I couldn't say 100% I didn't terminate.

This has happened many times over the years.  I wonder if there's a more robust way to protect the relational data from corruption.  One of the ways I use relational data is to flag Favourite Artists (relation on Artist).  My entire set of tags was wiped.  It's unlikely I'd even be unflagging one artist, let alone all, making the "termination during updating" feel unlikely.

Just bringing it up because it has happened too many times and always a major headache to fix after corruption.

[mark_h]

And I've just lost a whole bunch of relational fields again...

Something definitely awry here.

Only major things I did yesterday were update MC and run through a big conversion job testing Hendrik's recent changes.  Just repeated those again and they didn't delete the relational fields, so really out of ideas.

But I do feel that maybe a catch, or even just a log entry, in the code that spots when relational fields are about to be wiped would be useful to help track down the problem.