EPICS Home

Experimental Physics and Industrial Control System


 
1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  <20052006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024  Index 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  <20052006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>

Subject: Re: mv162 and nfs
From: Steven Hartman <[email protected]>
To: Andrew Johnson <[email protected]>
Cc: Geoff Savage <[email protected]>, [email protected]
Date: Wed, 26 Oct 2005 17:52:47 -0400 (EDT)
On Wed, 26 Oct 2005, Andrew Johnson wrote:

> Have you run ifShow on both sides of a transfer and examined the
> differences in values?

And also nfsstat -s and netstat -i on the nfs server.

> Could this be anything to do with discarded packets?  NFS is prone to
> delays if packets are lost, and this might be the issue.

Yes.

> I don't know what the defaults are for these that are used when you
> supply zeros, but in Tornado 2.x we use 32 tfds and 128 rfds here -

Defaults are 32 and 32.

-- 
Steve Hartman
[email protected] || 919-660-2650
Duke Free Electron Laser Laboratory

References:
mv162 and nfs Geoff Savage
Re: mv162 and nfs Andrew Johnson

Navigate by Date:
Prev: motor record distribution drvGpib bug Ronald L. Sluiter
Next: Documentation update Hammonds, John P.
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  <20052006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: Re: mv162 and nfs Andrew Johnson
Next: motor record distribution drvGpib bug Ronald L. Sluiter
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  <20052006  2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024