From: Matt Triviski <matt.triviski@...>
Date: 2010-07-22T09:56:00+09:00
Subject: Re: Writing to file delay

Joel VanderWerf wrote:
> Eric Wong wrote:
>>> Redhat 5) is waiting until it gets a block of data before it actually
>> "file.sync = true" looks good already.
> You could try calling fsync as well (it's not a writer, just call
> "file.fsync"):
> 
> Implementation from IO
> ------------------------------------------------------------------------------
>    ios.fsync   -> 0 or nil
> 
> ------------------------------------------------------------------------------
> 
> Immediately writes all buffered data in ios to disk. Note that
> fsync differs from using IO#sync=. The latter ensures that data
> is flushed from Ruby's buffers, but doesn't not guarantee that the
> underlying
> operating system actually writes it to disk.
> 
> NotImplementedError is raised if the underlying operating system does
> not support fsync(2).


Hey fellas, thanks for the response.

I tried file.fsync, and checked to make sure that the return was not 
nil, and it wasn't.  However, I am seeing the same activity.  I took out 
the select call and as Eric suggested and just used 
"file.write(server.socket.readpartial(8k))" with and that did not seem 
to help either.

I am writing to NFS, so I am thinking that this may be the issue.  I'm 
starting to think that there is no way for me to get what I want, except 
to rewrite this in C or Assembly :(
-- 
Posted via http://www.ruby-forum.com/.