Re: oaf time stamp problem

["Troy D . Gillette" <tgillette turbolinux com>]

On Thu, 26 Oct 2000, Mathieu Lacage wrote:

> "Troy D . Gillette" <tgillette turbolinux com> writes:
> 
> > I'm not sure if this is the right place to post oaf problems, but...
> > 
> > impl_OAF_ObjectDirectory_get_active_servers inside oafd/od-corba.c uses
> > time stamps to decide whether to check active servers (specifically,
> > only_if_newer and servant->time_active_changed). However, since these are
> > only set to the nearest second, they are usually equal. At least in my test
> > case with only one server running. This causes any "_active" queries to fail
> > randomly. Changing the test from < to <= seems to fix the problem, but I'm
> > not sure how safe that is.
> 
> Are you pooling oaf to see if the server is activated ?

I'm sending "(_active == FALSE)" as the last part of my requirements to
oaf_activate(). But due to the time stamp problem, it fails 2 out of 3
times.


     Troy D. Gillette - Principal Software Developer - TurboLinux, Inc.
_____________________________________.oOo.____________________________________
Cabbage, n.:
A familiar kitchen-garden vegetable about as large and wise as a man's head.
    -- Ambrose Bierce, "The Devil's Dictionary"