[Asterisk-Users] UPDATE - 512 Calls w/ Dig Rec: NFS Setup and Benchmarks

[tim panton]

On 3 Oct 2005, at 22:54, Matt Roth wrote:

> List members,
>

It has been a while, but I once implemented a simple shared database  
over NFS, so
dredging my memory produced the following:

>
> Future Plans and Unresolved Issues
> ==================================
>
> I wrote Windows software for another project that mixes leg files,  
> indexes them by call time, and archives them after a given period  
> of time. I plan to port that code to a set of shell scripts that  
> will be run on the Digital Recording server out of cron. If anyone  
> knows of an existing project that has accomplished this already,  
> please let me know.
>
> Before writing these scripts, I have two questions that need answered:
>
> 1) How can I tell when a file is complete on the NFS server?

If I recall right, you can't (not on the nfs server end). The way I  
used to handle this was to
have the creating client rename the file once it has finished with  
it. (remove a leading dot
is good). I think you can assume (with a decent NFS implementation)  
that the rename won't
happen untill all the queued writes+close have occurred.

>
> 2) What will happen on the NFS client if the NFS server crashes (I  
> expect the leg files to be written to the local mount point until  
> the mount is reesablished)?

Nothing so tidy, certainly not on files that were open at the time of  
the crash. To get that behavior
even for new files you would need to un-mount the nfs filesystem on  
the client whenever there is a
crash. (Hmm, kinda like the translucent filesystem...)

Tim.