> Modern udev shouldn't be too much of a problem. However, programs with
> suboptimal implementation regarding the handling of many devices may
> slow down the system even worse if they see more events. 

If configuration events like the ones we're discussion happen often
enough to trigger this problem, I'd say the sub-optimal application is
the one that's clearly in need of remedy, not the system itself. 

> > In the case of FCP disks, I know *I* care, and so I
> > could decide what to do based on what happened.
> Hm, maybe I should reconsider... change events may not be that bad.

I think it would be a nice step toward the ability for the system to
react to new devices in a automated way. 

> Well, if your configuration change makes your system unusable, it is
> serious... 
> Most of the problem comes from the actions triggered by the
> events. Good luck trying to get the consumers fix their stuff :()

Your gun, your foot. 8-)

> > I guess it comes down to how frequently one reconfigures devices.
> Unexpected failures (which would be easily recoverable) may trigger an
> event storm as well...

I'm not sure there's much one can do about unanticipated failures --
those are going to be messy regardless. In a large configuration, that's
going to look like a large-scale reconfig, in which we have to just
power through the events and deal with it as queued. If the user configs
a heavyweight reaction, that's going to be what will be, but at least
he'll be *able* to configure the response. 

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to