On Mon, 16 Mar 2009, Brad Knowles wrote:

> on 3/16/09 11:34 AM, [email protected] said:
>
>> so it sounds like you are advocating configuring all the clients to only 
>> point at a single server, and if that server is down for any reason just do 
>> without NTP altogeather.
>> 
>> how is that a sane configuration?
>
> Instead of making silly assumptions out of ignorance and publicly 
> embarrassing yourself, you really should try actually reading the page at 
> <http://support.ntp.org/bin/view/Support/SelectingOffsiteNTPServers>.

who's talking about offsite ntp servers?

I'm thinking of the case where I have a pair of GPS based ntp servers on 
an internal network and pairs of firewalls seperating layers of networks.

I can either point my clients at the two proxy firewalls/ntp servers, 
point them at one (and not have any redundancy), point them at a VIP 
(which is evil according to you, and common ntp software won't do), or try 
and convince management that I need to add 50% more devices to solve some 
sort of critical NTP problem that hasn't manifested itself in the last 
decade that these systems have been running (and then if a device fails 
I'm back to two ntp servers, which is somehow total evil)

you are making the blanket statement that pointing at two ntp servers is 
worse than pointing at one. that may be from the ntp accuracy point of 
view (messing up the time by a few ms), but from a datacenter point of 
view where things need to keep running (and it's actually more important 
for all the systems to have the same time than it is for that time to be 
exactly accurate) you are dead wrong.

David Lang
_______________________________________________
Tech mailing list
[email protected]
http://lopsa.org/cgi-bin/mailman/listinfo/tech
This list provided by the League of Professional System Administrators
 http://lopsa.org/

Reply via email to