obsidian-dataview
obsidian-dataview copied to clipboard
blacksmithgu
lots of errors
What happened?
For a few weeks now... maybe since I upgraded to 0.5.x sometimes I get thousands of errors when starting Obsidian. Not always, but every now and then. I am guessing this has to do with something in DataView indexing... though I am not really sure.
do these errors look familiar?
DQL
No response
JS
No response
Dataview Version
0.5.34
Obsidian Version
0.15.1
OS
Windows
Those are indeed being caused by the Dataview indexing - Dataview uses IndexedDB, which is a local file cache that browsers generally provide by default.
Do you have any custom settings / browser configuration / plugins etc that disallow using local storage, or maybe some kind of security auditing that does this? Alternatively, perhaps there isn't any available storage space.
It could be storage related because it happens intermittently. If it were a custom setting to disallow local storage or a security thing, it would happen consistently every time. It has occurred both on my laptop and desktop a couple of times during the past few weeks.
from an obsidian perspective it seems I don't have an issue with local storage quota:
Looking around I did find that I was low on local storage for my system drive, I did a bit of cleaning up, and will see if this resolves the problem. That said, there is no space issue on my system drive on the laptop.
@blacksmithgu - so this error seems to happen when I am testing something and turn on and off Obsidian fast with a large test vault. I think the reload of the index does not finish before I close Obsidian, and when I start again some data is corrupted leading to this problem.
The good news is, that this is not a typical usage scenario. Also, this might give you a hint on how to reproduce this flood of errors.
I added a method for dropping the entire file index (Dataview > Force Drop Cache) - does that cause the issues to go away temporarily, out of curiosity? I may be able to recover from this error by doing that specifically to at least reduce thousands of errors to one recoverable error.