• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

NTP Services.

Keiichi

[H]ard|Gawd
Joined
Jun 10, 2004
Messages
1,491
I just wanted to get some thoughts about using a dedicated NTP appliance as opposed to just getting a server get the time off the internet. I am planning on using this device on my company's network to synchronize PCs, employee timeclocks, as well as regular clocks.
 
If you can find a place to put an antenna, using GPS for the time source might be a lot cheaper. I think the linux docs talk about how to configure NTP to use a gps device as the reference.
 
GPS is the way to go, but as has been said you'll need an external antenna. This also introduces a single point of failure - if that dies you're back to your servers local clock.

IMHO the only reason not to use Internet time sources is that you don't have Internet access from your server or you require an extremely accurate time source for, say, scientific purposes or machine control.

I don't know about Windows, but the reference implementation of NTP (http://www.ntp.org) allows for multiple sources. If you pick your time sources well (i.e. close to you in terms of round trip time, stratum 1 if you have a lot of hosts, stratum 2 if it's only a few 100) you should get .1 to .01s accuracy. You're only points of failure there would be the server and your internet connection which I'm sure would cause enough of a problem for people to notice and put the effort into fixing.
 
Yeah, echoing snowbee. There's no reason not to use your DCs as your NTP servers in your domain. Have the PDC sync with time.nist.gov or the like and you're set. You shouldn't get even a second off the official time, especially if you tweak it to sync every hour. The only reasons to use a different source would be no internet access or scientific applications.
 
actually, I wouldn't suggest syncing to time.nist.gov, I'm sure they get enough traffic as is. us.pool.ntp.org should give you a good selection of servers to use that should still be plenty accurate.
 
We have a GPS timeclock where I work, technically it's a BITS clock that connects via a DS0 into our softswitch to provide a timing source to the switch which in turn provides timing to all other circuits connecting to the softswitch (DS0, DS1, DS3's...etc). The BITS clock has a license for an NTP server, the company threw it in when we purchased ours. While internet time would work fine for all of my servers it's nice to have a highly accurate Stratum 2 server to pull from, however, over time all of our equipment builds a drift reference so even if no atomic clock is available the systems would do fine for a few days without any really noticeable drift.
 
I'm basically seeing that getting a dedicated gps device is probably going to be a waste. Unless they make em for under $500 or so.
 
I've had very good luck synching to the ntp.org pool.

You could do this on your server and/or workstations:

net time /setsntp:us.pool.ntp.org

That will set the machine to synch to the ntp.org pool.
 
Back
Top