View Single Post
Old 08-19-2015, 07:57 AM   #61
knc1
Going Viral
knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.knc1 ought to be getting tired of karma fortunes by now.
 
knc1's Avatar
 
Posts: 17,212
Karma: 18210809
Join Date: Feb 2012
Location: Central Texas
Device: No K1, PW2, KV, KOA
Quote:
Originally Posted by NiLuJe View Post
Try putting it in base-us instead, us is a weird crappy fuse proxy.

That, and a 1MB file might be too small.
Yup, I seem to recall twobob/geekmaster running into that holy message.
Plus up the size, that 1M probably would not even hold the swap metadata.
You have gigabytes of space, try something useful, like: count=64

- - - -

In older (2.6.x) kernels, there was a possibility of a deadlock or race condition in the:
swap -> loop-device -> file -> file system -> device
pathway (one or more of those layers can end up generating another swap request {when VFS needs more buffers, IIRC}).

Plus, as NiLuJe points out, there is a funky fuse proxy already in the /mnt/us path.
Try to get under the proxy using base-us instead.

Last edited by knc1; 08-19-2015 at 08:05 AM.
knc1 is offline   Reply With Quote