does 'noatime' make sense for a read-only filesystem?

mount -o remount,ro,noatime

As the title says - if the filesystem is actually read-only, does noatime have any affect?

gng

Posted 2011-10-20T11:07:53.517

Reputation: 31

Answers

No, read only means read only, if you told it to not write to it it will not write access times to it.

For more information, see man fstab.

ktf

Posted 2011-10-20T11:07:53.517

Reputation: 2 168

ext4 replays the journal even on readonly mounts. So a readonly mount will still write to the underlying block device. (whether it processes atime is another matter). – sourcejedi – 2017-10-18T16:51:00.540

I see, if you do not mind me asking - source on that? – gng – 2011-10-20T11:19:26.870

My experience :-) No, really: ro means NO writes at all. Everything else may safely considered as bug. – ktf – 2011-10-20T11:23:00.873

I'm not sure about that. It might that VFS handles noatime somehow different. A little bit of google search shows lots of people using both at the same time, but I couldn't get any source about why doing it or not, neihter in the man pages. – Jens Erat – 2011-10-20T11:53:07.860

On my system the man page stated ro means "read-only" (phew, lucky me). Adding noatime to ro helps as much as finger-crossing: Will do no harm! – ktf – 2011-10-20T13:05:24.063

Short answer: No.

If the filesystem is mounted read-only, then the kernel module should be sufficiently smart to not even think about writing atime entries for accessed files, because that would constitute a ... write, which is verboten by the read-only status.

killermist

Posted 2011-10-20T11:07:53.517

Reputation: 1 886

Possibly. Depending on the filesystem.

Actually things here are much more interesting and odd than you might think.

...

Mounting your NFS filesystems with noatime will reduce the number of NFS requests you make to the server.

https://utcc.utoronto.ca/~cks/space/blog/linux/NFSReadonlyAtime

sourcejedi

Posted 2011-10-20T11:07:53.517

Reputation: 2 292