Peter Memishian writes:
> 
>  > Since other applications such as inetd and routed are all using /var/run, 
> wouldn't
>  > it be better to put all unix socket files in a unified place?
> 
> Assuming the file itself is an implementation detail (and I hope it is
> because otherwise it means applications have to directly code to a very
> low-level interface), it simply doesn't matter.  You should put the file
> where it gets the job done.

Perhaps I'm misunderstanding the issue here, but why would VRRP _need_
to be running so early in boot that we haven't yet mounted /var and
/var/run?

The point of VRRP is to provide address fail-over between distinct
systems.  I see no reason to participate in that process until the
system is demonstrably "up" and ready to handle traffic -- and that's
certainly not true during the early boot process.

In fact, I would expect that having VRRP come up too early is a
hazard.  It allows us to grab the address and direct traffic towards
ourselves before we even know if we're going to come up.

Could someone clue me in on the design issue that forces the use of
/etc/svc/volatile (and that would obviously also force the daemon and
supporting libraries into the root file system)?

-- 
James Carlson, Solaris Networking              <[EMAIL PROTECTED]>
Sun Microsystems / 35 Network Drive        71.232W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.496N   Fax +1 781 442 1677
_______________________________________________
networking-discuss mailing list
[email protected]

Reply via email to