I have an SSD and I’m trying to set noatime to save writes. I didn’t do this at the time of install, so I edited /etc/fstab later on. I noticed, however, that when I right-click on a file and view its properties, it still gives me a “modified” and “accessed” time. Did I do something wrong? Here is my fstab:
noatime prevents updates of the accesstime on an access… it does not
prevent it from being seen/present. The write (on every read) is
where performance is lost… not on reading. Thus that is the part which
is disabled with noatime.
A proper test is to read the file and then observe that the access time is
NOT changing.
Good luck.
On 11/09/2010 09:36 AM, dan8029 wrote:
>
> Hi everyone,
>
> I have an SSD and I’m trying to set noatime to save writes. I didn’t do
> this at the time of install, so I edited /etc/fstab later on. I noticed,
> however, that when I right-click on a file and view its properties, it
> still gives me a “modified” and “accessed” time. Did I do something
> wrong? Here is my fstab:
>
>
> Code:
> --------------------
>
> /dev/disk/by-id/ata-INTEL_SSDSA2M080G2GN_CVPO0161048A080JGN-part1 / reiserfs acl,user_xattr,noatime 1 1
> /dev/disk/by-id/ata-INTEL_SSDSA2M080G2GN_CVPO0161048A080JGN-part2 /home reiserfs acl,user_xattr,noatime 1 2
> /dev/disk/by-id/ata-INTEL_SSDSA2M080G2GN_CVPO0161048A080JGN-part4 /windows/C ntfs-3g users,gid=users,fmask=133,dmask=022,locale=en_US.UTF-8 0 0
> proc /proc proc defaults,noatime 0 0
> sysfs /sys sysfs noauto,noatime 0 0
> debugfs /sys/kernel/debug debugfs noauto,noatime 0 0
> usbfs /proc/bus/usb usbfs noauto,noatime 0 0
> devpts /dev/pts devpts mode=0620,gid=5,noatime 0 0
>
> tmpfs /var/log tmpfs defaults,noatime,mode=0755 0 0
> tmpfs /tmp tmpfs defaults,noatime,mode=1777 0 0
> tmpfs /var/tmp tmpfs defaults,noatime,mode=1777 0 0
>
> --------------------
>
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.15 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
Glad to hear it. Thank-you for posting back a confirmation.
Good luck.
On 11/09/2010 04:06 PM, dan8029 wrote:
>
> Oh, you’re right. The access time of my files is never changing. It must
> be set up right. Thanks for the help!
>
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.15 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
Ooh can I ask one more question? I’m using ReiserFS. I love Reiser. Should I be using EXT2 though, on a SSD? I thought I should have a journaling FS such as Reiser because this is a laptop, and sometimes batteries die and stuff.
ext2 is a fine filesystem, but it doesn’t have journaling so it’s not what
you’re after. ext3 has journaling but I’d personally go for ext4 or xfs
as ext4 is ext3++ and xfs is just pretty stinkin’ neat. I do not think
any single filesystem has insane perks over others when it comes to
SSD,but ext4 has some definite perks over ext3 and xfs is great in its own
right.
Good luck.
On 11/09/2010 06:36 PM, dan8029 wrote:
>
> Ooh can I ask one more question? I’m using ReiserFS. I love Reiser.
> Should I be using EXT2 though, on a SSD? I thought I should have a
> journaling FS such as Reiser because this is a laptop, and sometimes
> batteries die and stuff.
>
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.15 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
On 2010-11-10 03:01, ab@novell.com wrote:
> ext2 is a fine filesystem, but it doesn’t have journaling so it’s not what
> you’re after.
No journal could be a good thing on a flash drive, as the journal is
written often and in the same place, so it wears out. There should be
special devices that have the journal in battery backed RAM.
–
Cheers / Saludos,
Carlos E. R.
(from 11.2 x86_64 “Emerald” at Telcontar)
I’ve heard from several sources that the old concerns about writes with
journaling and swap are not justified, especially when compared with the
benefits that journals provide.
On 11/10/2010 06:39 AM, Carlos E. R. wrote:
> On 2010-11-10 03:01, ab@novell.com wrote:
>> ext2 is a fine filesystem, but it doesn’t have journaling so it’s not what
>> you’re after.
>
> No journal could be a good thing on a flash drive, as the journal is
> written often and in the same place, so it wears out. There should be
> special devices that have the journal in battery backed RAM.
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.15 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
Assuming that it is an SSD, and not some other kind of flash device, a journal shouldn’t really be a problem with newer devices. The controller does some level of remapping, and that limits the ‘re-writes to one physical location’ at the level of the actual flash device. Well, that’s what the experts say.
BTRFS has a special SSD mode of some kind, but BTRFS is a bit new and I’m sure that a minority interest option like SSD mode has received much less testing than ‘plain’ BTRFS, I’d say it was still probably a bit soon. And, the last results that I saw weren’t really much better, in terms of speed, than plain-BTRFS mode. Now speed isn’t really the main issue here, but write amplification/remapping is, but I get the impression that the SSD mode doesn’t do much yet (I think that it may only adjust block sizes, which might mean that it will never do anything in read speeds, but be useful in terms of writes).
If I had to do it today, I’d go with EXT4, but next year that might be BTRFS, I don’t know.
The write (on every read) is