[ale] USB->IDE problems

[Christopher Fowler]

That is a good idea. I would have to keep flash size limited to what I
can spare in ram
 
On Thu, 2004-01-22 at 10:18, Dow Hurst wrote:
> How are hotswap hard drives taken care of?  Can you treat 
> the USB device like it is part of a RAID array and is an 
> extra mirror that is hotswappable?  You'd need a RAM drive 
> as the primary drive and some RAID code for the USB device.
> Dow
> 
> 
> Chris Fowler wrote:
> > I would like to write log data to an external thumb drive.  If the user
> > unplugs it to read it then I need to be able to stop writing.  When they
> > plug it back in I start writing data again.  I do not intend on
> > buffering data until they plug back in.  If it does not exist then the
> > data would not be written.
> > 
> > I've rewrote much of the hotplug code in C because the hotplug stuff
> > that is on kernel.org and in RH is written in bash.  I can not really do
> > IPC with bash.  
> > 
> > I can have hotplug start up the streamers when it is plugged in.  Then
> > the streamers will be terminated when it is unplugged.  It is also
> > possible that the streamers could always be running.  They would not
> > carry file descriptors around for the log files on the flash.  They
> > would open(), write(), close().  Then go into a select() loop waiting on
> > data to write to flash.
> > 
> > 
> > On Wed, 2004-01-21 at 21:17, Geoffrey wrote:
> > 
> >>Chris Fowler wrote:
> >>
> >>>There may be some corruption of the file in vfat.  But that is not
> >>>really of a concern.  I can recover from that in my code.
> >>
> >>So I guess I don't understand the problem.  I'd suggest you'd be in the 
> >>same boat with ext2, ext3, Reiser or whatever.
> > 
> > 
> > _______________________________________________
> > Ale mailing list
> > Ale at ale.org
> > http://www.ale.org/mailman/listinfo/ale
> >