testing performance

[Gary E. Miller]

Yo Hal!

On Tue, 06 Aug 2024 17:48:11 -0700
Hal Murray via devel <devel at ntpsec.org> wrote:

> [From https://gitlab.com/NTPsec/ntpsec/-/merge_requests/1399]
> 
> Gary said:
> 
> > But I agree with you that howto run non-root needs to be
> > documented, and I would also like tests in ntpd to verify the
> > needed CAPS to run as designed.  
> 
> I expect it will crash if it doesn't have the appropriate caps.
> Maybe blunder on after an error message if it can't open something
> needed by a refclock.

Add an option, people will use it.  Since this one will be hard to use
properly, the documentation and error messages need to be very good.

> > Then we can test how bad the performance really is.   
> 
> What do you have in mind?

I would just look at jitter.  I expect it to be very bad.

> I'm scheming to setup a chain of servers to see if the PLLs are
> stable.

Servers, hosting containers of serrvers.  With ntpd running in the server
in the containr.  I have no idea how to tell the container to allow
adj_time() to work.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
	gem at rellim.com  Tel:+1 541 382 8588

	    Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin