Messages in this thread | | | Date | Sun, 15 Aug 1999 16:42:24 -0600 | From | Richard Gooch <> | Subject | Re: kupdate & laptop's [patch for integration of mobile-update] |
| |
Zack Weinberg writes: > Richard Gooch wrote: > > Peter Englmaier writes: > > > I tried to deactivate everything, but even in single user mode and > > > no process or daemon running, I get disk I/O. Yes it stops after a > > > while, but a simple 'ls' triggers new activity even if all is in > > > cache already (ls finished and a second later the disk i/o happens). > > > Perhaps this is some timestamp business, and how can I hope to patch > > > every user land command to avoid it? With mobile-update you can > > > edit, compile, run programs without any disk i/o. > > > > There must be something wrong in your setup. I don't experience these > > problems. If a simple <ls> triggers a spin-up, then check to make sure > > you're mounting with the "noatime" flag. By default, a simple read of > > a file/directory will cause the inode access time to be updated, which > > in turn requires the disc to be written to. > > Noatime helps but is not a panacea. Even with noatime set, an > opendir/readdir/closedir sequence will spin up the disk every time. > I have no idea why. I don't really mind ls spinning it up - if I'm > actually working on the machine, it's fine for the disk to be > running. I just want it off when the machine is idle, because it > generates a ridiculous amount of heat when it's spinning.
Must be a configuration problem. If I do <ls -lF /bin> then my disc does *not* spin up, once my cache is warm.
> I'd like to see atime updates be recorded in memory but not flushed > back until the inode cache is shrunk, the fs is unmounted, or we > need to write back the inode for some other reason. Sort of like > the current handling of dtime in ext2. This should give the speed > improvement without losing the information.
Except a big recursive diff can then lead to a delayed storm of writes. Very annoying. One of the reasons I mount with noatime.
Regards,
Richard.... Old: rgooch@atnf.csiro.au Current: rgooch@ras.ucalgary.ca
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |