[sheepdog-users] Storage limit to 4T per node with full replica and less with erasure code

Valerio Pachera sirio81 at gmail.com
Tue Sep 16 15:20:16 CEST 2014


2014-09-16 12:50 GMT+02:00 Ruoyu <liangry at ucweb.com>:
> What is the sheep's version?
> There is a bug about fd leak in sheep daemon if the version is older than
> 0.8.3

Sheepdog daemon version 0.8.2

> Please see the detail about the patch:
> https://github.com/sheepdog/sheepdog/commit/580b072a5bae4b665b784fe4cb93a21f05c1ed63.
>
> If not applying the patch, even if you set nofile of your system up to
> 6144000, the problem will be occurred in some day of the future.

> It is not correct. The fd number is irrelevant to the object size.
> If your data objects do not be read and written at the same time, the fd
> number to be consumed is not high.

I was considering df in a wrong way.
I guess is a sort of index of the opened file.
When sheepdog access an object, it creates a file descriptor.
It should remove the file descriptor once stop accessing the object.
The default value in debian is (Hn 4096 Sn 1024).
This means my host reached the 4096 limit.

Now I have set the limit to 1024000, so I can relax.

> If not applying the patch, even if you set nofile of your system up to 6144000,

How can I monitor the number of fd?
(Just to understand if 0.8.2 needs the be patched).



More information about the sheepdog-users mailing list