[plug] handling failed non-redundant storage in a server

James Devenish devenish at guild.uwa.edu.au
Thu Feb 12 11:57:19 WST 2004


In message <1076557732.19461.12.camel at bucket.localnet>
on Thu, Feb 12, 2004 at 11:48:52AM +0800, Craig Ringer wrote:
> >  Would prefer to have
> > programmes die "cleanly" with 'file not found' than hang.
> 
> There are times when it's appropriate for a program to hang. For one
> thing, the kernel doesn't know if the device will be coming back

Oh yes, of course. I would always expect some form of timeout or
administrative action require first. Giving the benefit of the doubt, I
suppose that 'uninterruptabe processes' are possibly an 'acceptable'
side effect or maybe even an intended consequence of some otherwise
'really good' part of the kernel. I suppose it does seem that
uninterruptable processes are usually (but not always) associated
with obvious I/O problems (or 'too new' code in the kernel).





More information about the plug mailing list