Reindex Often / Over 10 hours per

Discussion related to "Everything" 1.5 Alpha.
Post Reply
MrKuenning
Posts: 3
Joined: Mon Jan 22, 2024 11:04 pm

Reindex Often / Over 10 hours per

Post by MrKuenning »

First of all, I love everything, it's one of my favorite and most used apps on my PC. I use it as a photo gallery, file manipulator... The use cases never end.

Recently, I set it up to index People and Tags from images. This scan takes about 12 hours to run, as it has to parse through over 3 million images across my PC and my NAS. I don't mind that it takes this long. But I do mind that it will just randomly decide it needs to reindex. When it starts a reindex. I can no longer use any of the indexed metadata for 12 hours until it finishes again.

I have Everything 1.5 indexing 8 drives on my PC, 4 of them come and go.
I also have everything indexing 15 network UNC paths via Folders.

I have the option to remove offline volumes unchecked and I have

Code: Select all

keep_missing_indexes=1 
set in the INI

It will work fine and then randomly start indexing for 12 hours again.
When I check the statistics it says
Last rebuild reason: Out of date NTFS volume E: is available
But nothing changed in my E: drive that I am aware of.

Is there any other ways to tell it to not do a full reindex?
Would I be better off setting up an everything server?
void
Developer
Posts: 15506
Joined: Fri Oct 16, 2009 11:31 pm

Re: Reindex Often / Over 10 hours per

Post by void »

Thank you for your feedback MrKuenning,



Everything should only index your property values once.

Everything will rescan your properties if the volume is reindexed.
This can occur if you change your indexing settings or if the volume Journal ID changes.



I am working on making Everything reuse the existing property values.
This means you can search and display your existing properties immediately after a reindex.
However, Everything will still need to rescan all these properties.



Last rebuild reason: Out of date NTFS volume E: is available
The USN Journal on your E: drive was deleted or too many changes occurred on your E: drive while Everything was not running.



A more robust property server is on my TODO list for Everything 1.6.
Post Reply