Re: NTP 4.2.6p3 multicast server not visible in orphan peer to peer mode?

On 4/3/2011 6:30 PM, Wu, Bailey wrote:
I have an isolated ntp network with 2 servers running
ntp 4.2.6p3. Clients run various other version of ntp.
My problem is to do with the servers and 2 HPUX multicast client.
The 2 servers are setup in orphan mode and are peer'd
with each other. While running server 1 will eventually
synchronise with server 2, which I guess must have the
better clock.
In addition the servers are BOTH configured as multicast
servers to output via two network interfaces.
The HPUX clients are connected to both the servers
(all interfaces), so I would think the HPUX clients
would see 4 sources of multicast, but they do not.
If server 1 synchronises with server 2 then the HPUX
clients will only see 2 server 1 sources and vice
versa if server 2 is sync'd with server 1.
This means when one of the servers goes down the HPUX
clients are not able to select another source to
synchronise with.

My question is whether ntp in orphan peer to peer mode will automatically disable one of the servers multicast if it is the primary server? The HPUX clients have ntp 3.5f.

I'm running 4.2.7p128 on a doz M$ Win and a few Linux.
They all see & use each other each other via manycast,
{even more so when they are routinely isolated from the internet,
(but not each other, for the most part)}.

Config(s) {at the moment, in general}:

# ---------|---------|---------| Organization NTP Clients
setvar access_policy = "NTP Manycast / Multicast LAN Client" default

statsdir /var/log/ntpstats/
driftfile /var/lib/ntp/ntp.drift
keys /etc/ntp/ntp.keys
trustedkey 1

restrict -4 default limited kod nomodify notrap
restrict 127.0.0.1
restrict -6 default limited kod nomodify notrap
restrict ::1
restrict 224.0.1.1 mask 255.255.255.255 nomodify
restrict source nomodify

tos floor 2 cohort 1 orphan 11

manycastserver 224.0.1.1
manycastclient 224.0.1.1 key 1 preempt prefer
multicastclient 224.0.1.1 key 1 preempt prefer
broadcastclient

# Organization NTP Servers
server margo.anitech-systems.net key 1 iburst preempt prefer minpoll 9 # 8min
....

# UpStream ISP Stratum 2 Servers
server ntpB.isp.invalid iburst minpoll 11 # 30min

# OS Vendor Servers
server ntp.os.invalid iburst minpoll 12 # 1hr

# Regional NTP Pool Servers
server us.pool.ntp.org iburst preempt minpoll 10 # 15min




# ---------| Organization NTP Servers, add to the above client config
setvar access_policy = "via pool.ntp.org / Multicast Server" default

tos minsane 4 minclock 6 maxclock 20 floor 1 cohort 1 orphan 11

# Other Stratum 2 Servers
server ntp1.other.invalid iburst minpoll 10 # 15min
server ntp2.other.invalid iburst minpoll 10 # 15min

# Other Stratum 1 Servers
server ntpA.isp.invalid iburst minpoll 10 # 15min
server ntp.nist.invalid iburst minpoll 10 # 15min





I intentionally didn't use pool e.g
pool us.pool.ntp.org iburst preempt minpoll 10 # 15min
as I didn't want them spinning up more than 1 from the pool,
rather them filling in up to maxclock with manycasts.

--
E-Mail Sent to this address <BlackList@xxxxxxxxxxxxxxxxxxx>
will be added to the BlackLists.
.

Relevant Pages

  • Re: ntp server pool advice
    ... this ntp network should give time to specific clients devices and not anyone on the public network. ... how many ntp servers should be needed to handle that much request knowing that each card handle 10,000 request per sec? ... A would not use load balancing for NTP servers. ...
    (comp.protocols.time.ntp)
  • Proposed NTP solution for a network
    ... Locations A and B have a large community of Suse 10.x Enterprise servers, each with very stringent requirements to have time be very closely "in sync" with each other at that site, as well as at the other site. ... Each of A, B, and C have a dedicated NTP appliance, with integrated GPS receiver and antenna on the roof. ... This would enable the appliance to lose the reference source and still be useful to the Stratum 2 servers that are clients of these appliances. ...
    (comp.protocols.time.ntp)
  • Re: ntp server pool advice
    ... this ntp network should give time to specific clients ... according to your advice, better not using load balancer, thats good ... First, each client should have at least 4 configured servers, so you can use the same ntp.conf file for all of them. ... Second, if you really can handle 10K requests/second per card, then that means that you can handle 640K clients per card, with worst-case polling. ...
    (comp.protocols.time.ntp)
  • Re: Hardware SNTP server
    ... But I certainly see what appears to be a pretty big drawback in the implementation, in that you're throwing away most of what I consider to be the best parts of NTP, just so that you can burn these things into FPGAs. ... Because these servers would not be the same and would not have precisely the same concept of time, to do this would be to completely destroy the NTP protocol itself. ... Even if they're all served by the same atomic clock reference, using serial cable splitters or whatever, those differences in cable lengths would be significant. ... What you want is broadcast, multicast, and manycast servers and clients, so that the load is automatically distributed across the set of servers, and the clients are automatically served by those machines which are close by. ...
    (comp.protocols.time.ntp)
  • Re: Dropped Connection to Shared Drive
    ... I believe that is the latest version and feature pack. ... and NTP features on servers. ... All my clients report to the SBS server for updates. ... However I do not run the NTP or the PTP on my servers as recommended by ...
    (microsoft.public.windows.server.sbs)