ERROR : Error traversing cache ... Access is Denied

[klunky]

@asdffdsa

• is there a specific reason you are using the cache?

No... i am just using it ... i should disable to test ? If yes which parameters ?

• the w:\ drive, is it local or network

iSCSI (both S:\ and W:\ from Synology)

Answered above.

[asdffdsa (jojothehumanmonkey)]

well, i suggested how to test this above, by removing as many flags as possible, test and add flags back in a certain order and keep testing.

the cache backend is beta.
i could be wrong but it has known bugs and nobody is maintaining the source code.
read this:
https://rclone.org/cache/#windows-support-experimental
"There are a couple of issues with Windows mount"
"It should be considered as experimental for this OS."

[calisro (Rob)]

@klunky

i am directly seeding from gdrive mount.

That sounds like a ban waiting to happen. I'd suspect you'll quickly see 403s.

[klunky]

I think I have found the problem, even with cache disabled I get errors, it has to do with the PATH file length of --cache-dir

I will do some tests tomorrow, to be sure that this is the issue.

Current mount setup:

rclone.exe mount drive: C:\gdrive ^
--allow-other ^
--cache-db-purge ^
--buffer-size=512M ^
--dir-cache-time=24h ^
--drive-chunk-size=512M ^
--fast-list ^
--use-mmap ^
--vfs-read-chunk-size=512M ^
--vfs-read-chunk-size-limit=off ^
--vfs-cache-mode=writes ^
--vfs-cache-max-age=360m ^
--vfs-cache-poll-interval=1m ^
--cache-dir="W:\z" ^
--transfers=40 ^
--checkers=40 ^
--tpslimit=10 ^
--stats=1m ^
--stats-log-level=DEBUG ^
--drive-acknowledge-abuse

[klunky]

@calisro

That sounds like a ban waiting to happen. I'd suspect you'll quickly see 403s.

Sometimes that happens but it's rare... do you recommend any different parameters to maximize performance ?

[thestigma]

That config kind of looks like a mess to me... several flags there will do nothing at all in this context, there are random duplicates of the same flags and just in general it makes me not sure if your setup is intentional with knowing what all of these do or if it's just a result of picking up random "tips" from many different places.

I am not trying to offend you (and maybe you are very experienced and set this exactly as you intended for all I know, but this is jus the impression I get). I am just saying - less if often more, and unless you are really sure you understand the flags do and know why you need them, it's generally a good idea to leave them default.

Are you actually using the cache backend (remote)? I want to make sure you are not confusing the VFS/mount and the cache backend - which are completely different systems (you have flags for both).

Also in general I don't think I would recommend using the tmp-upload feature of the cache-backend. I saw way too many bugs in this in my testing, and unfortunately the cache backend remote is basically abandoned (quite a while ago) so these bugs are unlikely to be fixed.
Some parts of the cache backend actually do work fairly well, but the tmp-upload feature is not one of them (in my experienced opinion).

I can not pinpoint the error here more than to say it is likely happening in the VFS cache, or in some interaction between the VFS cache and the cache-backend. The convoluted config and use of tmp-upload function (which I don't even want to try to troubleshoot for honestly) makes this hard to diagnose just from this information.

TLDR - I would:

• Simplify the config for now to strictly what you know you absolutely need
• Disable tmp-upload for now
• Retest

Then if all is well you can maybe start adding back some of the less critical flags that you still want - and if you are not 100% sure of what they really do then just ask. It is better to ask than to indiscriminately copypaste something some guy said on the forum (which even if correct may be inapplicable for your setup or use-case).

[thestigma]

@klunky

Sometimes that happens but it's rare... do you recommend any different parameters to maximize performance ?

Seeding torrents via a clouddrive mount is never going to be "optimal" with any settings as there is just too much random access, and the latency on requests is so (relatively) high.
But if you wanted to optimize as much as you could you probably want to do the opposite of what you have done here... relatively small VFS read-chunks and definitely much much less buffer-size ect.

I will be happy to give you specifics and generally go though the flags with to make sure you get what you actually want out of it - but let's deal with one problem at a time here ok? I feel like rewamping your config is going to be like a whole different topic and best handled after your main problem.

[klunky]

Sorry for just getting back to this thread, had some problems on RL but I am back now,

@thestigma thanks for your update, let's fix each thing first

I think I am able to reproduce the error of VFS (characters length). Will try to do it in the next few hours and update the thread with the evidence