Rich Freeman on 28 Sep 2017 12:42:50 -0700


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] Reclaiming inodes


On Thu, Sep 28, 2017 at 3:28 PM,  <bergman@merctech.com> wrote:
> In the message dated: Thu, 28 Sep 2017 15:13:37 -0400,
> The pithy ruminations from Rich Freeman on
> <Re: [PLUG] Reclaiming inodes> were:
> => On Thu, Sep 28, 2017 at 3:07 PM, Steve Litt <slitt@troubleshooters.com> wrote:
> => >
> => > OK, I believe you. So let me rephrase:
> =>
> => Sorry if that came across as confrontational.
> =>
> => >
> => > If you have an SSD, frequently run fstrim on all your SSD partitions to
> => > reclaim deleted space. Doing so has many benefits.
>
> Um, depends on your definition of "frequently". According to the fstrim man page:
>
>        Running fstrim frequently, or even using mount -o discard, might
>        negatively affect the lifetime of poor-quality SSD devices.
>        For most desktop and server systems a  sufficient  trimming
>        frequency is once a week.  Note that not all devices support a
>        queued trim, so each trim command incurs a performance penalty
>        on whatever else might be trying to use the disk at the time.
>
>         From: util-linux-2.28.2-2.fc25.x86_64

True, though this is more a problem with brain-dead drives than fstrim
itself.  There is no reason that fstrim ought to reduce drive life and
if anything it should improve drive life by reducing copying of unused
blocks.  However, if the drive is brain-dead then of course anything
is possible.

> =>
> => fstrim has a lot of benefits (which I'm sure google can lead anybody
> => to).  It just doesn't have anything to do with reclaiming inodes.  As
> => far as I'm aware inodes go away as soon as you unlink anything in all
>
> Nope.
>
> [pseudo-code]
>         int fd;
>         fd = open(filename, O_RDWR);
>         unlink(filename);
> [/pseudo-code]

Yeah, I wasn't really thinking about reference counting here.  Ditto
for removing a hard link when other hard-links are present.

-- 
Rich
___________________________________________________________________________
Philadelphia Linux Users Group         --        http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion  --   http://lists.phillylinux.org/mailman/listinfo/plug