Hi Josef,
On Thu, Sep 18, 2014 at 11:20:02AM -0400, Josef Bacik wrote:
> --- a/fs/btrfs/disk-io.c
> +++ b/fs/btrfs/disk-io.c
> @@ -1764,6 +1764,7 @@ static int cleaner_kthread(void *arg)
> }
>
> btrfs_run_delayed_iputs(root);
> + btrfs_delete_unused_bgs(ro
On Thu, Sep 18, 2014 at 11:20:02AM -0400, Josef Bacik wrote:
> This patch has the side effect of no
> longer having a bunch of BUG_ON()'s in the chunk delete code, which will be
> helpful for both this and relocate. Thanks,
Would be better to split the unrelated error handling changes to a
separa
One problem that has plagued us is that a user will use up all of his space with
data, remove a bunch of that data, and then try to create a bunch of small files
and run out of space. This happens because all the chunks were allocated for
data since the metadata requirements were so low. But now
One problem that has plagued us is that a user will use up all of his space with
data, remove a bunch of that data, and then try to create a bunch of small files
and run out of space. This happens because all the chunks were allocated for
data since the metadata requirements were so low. But now
On 09/14/2014 12:46 AM, Wang Shilong wrote:
Hi Josef,
One problem that has plagued us is that a user will use up all of his space with
data, remove a bunch of that data, and then try to create a bunch of small files
and run out of space. This happens because all the chunks were allocated for
d
Hi Josef,
> One problem that has plagued us is that a user will use up all of his space
> with
> data, remove a bunch of that data, and then try to create a bunch of small
> files
> and run out of space. This happens because all the chunks were allocated for
> data since the metadata requiremen
On 09/12/2014 03:18 PM, Josef Bacik wrote:
> One problem that has plagued us is that a user will use up all of his space
> with
> data, remove a bunch of that data, and then try to create a bunch of small
> files
> and run out of space. This happens because all the chunks were allocated for
>
One problem that has plagued us is that a user will use up all of his space with
data, remove a bunch of that data, and then try to create a bunch of small files
and run out of space. This happens because all the chunks were allocated for
data since the metadata requirements were so low. But now
