Same fs instance everywhere

[hsablonniere]

Hello @wmhilton and team :wink:

One liner

I think the docs should make it very important that your app needs to have one and only one instance of new FS('fs').

Details

We only tested on Firefox, but in one of our codebase we had several instances of fs (in different modules) with const fs = new FS('fs'). We hoped they would be in sync since it's the same IndexedDB behind. It seems like there is a cache layer or something with the transactions. Maybe it comes from how idb-keyval works.

Anyway, if you don't use the same fs instance when you try to readdir or any other fs operations than the instance you used with git.plugins.set('fs', fs); you will have sync problems.

I know the getting started for browser mentions a global var but it does not say much more.

Questions

• Is my diagnose wrong?
• What do you think about this?
• Should we improve the docs?
• Should we find a way for several new FS('fs') to reuse the same wrapper behind the scenes?

Thanks :smile:

[billiegoose]

• Is my diagnose wrong? Nope! That's correct, if you have create multiple instances you'll have multiple filesystems. Which is fine if you do something like

const fs1 = new FS('global')
const fs2 = new FS('third-party')

but indeed would have very strange effects if the two instances are using the same IndexedDB backend! like

const fs1 = new FS('global')
const fs2 = new FS('global')

they'd end up fighting and clobbering each other. :(

• What do you think about this?

I think it's a footgun!

• Should we improve the docs?

That's probably the quickest solution.

• Should we find a way for several new FS('fs') to reuse the same wrapper behind the scenes?

It would be nice if the API made such mistakes impossible though... but if we kept references that would make filesystems impossible to garbage collect... unless we used WeakMaps but then I'm not sure whether that's supported in all the browsers lightning-fs currently works in......

So sorry you ran into this issue. Just in case you're wondering, you run into a similar issue if you try to access the same FS from the main thread and a worker thread (which has had me pondering how to make the fs thread-safe).

I'll try to update the documentation later tonight after my bike ride.

[billiegoose]

oh that's why this wasn't closed. I'd forgot to merge the PR 🤦‍♂

[isomorphic-git-bot]

:tada: This issue has been resolved in version 3.3.0 :tada:

The release is available on:

• npm package (@latest dist-tag)
• GitHub release

Your semantic-release bot :package::rocket:

[billiegoose]

@hsablonniere the situation should be improved now. The latest version of LightningFS is now threadsafe.

Using a single FS instance per thread is still better, because it won't have to compete for the mutex, but the two instances won't get out of sync anymore.