login
Header Space

 
 

Re: Understanding I/O behaviour - next try

Score:
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: <spamtrap@...>
Cc: Fengguang Wu <wfg@...>, <linux-kernel@...>, <mingo@...>
Date: Thursday, September 13, 2007 - 10:17 am

On Wed, 2007-08-29 at 01:15 -0700, Martin Knoblauch wrote:

g,

NFS has the concept of unstable storage, that is a state where it is
agreed the page has been transferred to the remote server, but has not
yet been written to disk.


see: /proc/sys/fs/nfs/nfs_congestion_kb

That is the limit for when the nfs BDI is marked congested, so
nfs_writeout + nfs_unstable <=3D nfs_congestion_kb

The nfs_dirty always being 0 just means that pages very quickly start
their writeout cycle.
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
Understanding I/O behaviour - next try, Martin Knoblauch, (Tue Aug 28, 11:53 am)
Re: Understanding I/O behaviour - next try, Chuck Ebbert, (Wed Aug 29, 12:25 pm)
Re: Understanding I/O behaviour - next try, Martin Knoblauch, (Wed Aug 29, 5:43 pm)
Re: Understanding I/O behaviour - next try, Jens Axboe, (Wed Aug 29, 5:48 am)
Re: Understanding I/O behaviour - next try, Martin Knoblauch, (Thu Aug 30, 6:50 am)
Re: Understanding I/O behaviour - next try, Martin Knoblauch, (Wed Aug 29, 10:26 am)
Re: Understanding I/O behaviour - next try, Fengguang Wu, (Tue Aug 28, 9:38 pm)
Re: Understanding I/O behaviour - next try, Martin Knoblauch, (Wed Aug 29, 4:15 am)
Re: Understanding I/O behaviour - next try, Peter Zijlstra, (Thu Sep 13, 10:17 am)
Re: Understanding I/O behaviour - next try, Fengguang Wu, (Wed Aug 29, 4:40 am)
Re: Understanding I/O behaviour - next try, Martin Knoblauch, (Wed Aug 29, 5:22 am)
speck-geostationary