Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

XFAC (?)

2,008 views
Skip to first unread message

Laurent Archambault

unread,
May 11, 2010, 1:58:54 PM5/11/10
to
After many search, i has not find my answer, but for the human that
>the status "XFAC" represent (?). I has find this "Extra Fast Attack
>Crafts", and i am very happy for this...but in french language it's
>very difficult to understand this word "XFAC".

--
----~o00o-----//{ ´°`(_)´°` }\\-----o00o~------

Laurent Archambault
Under Linux
_______________________________________________
questions mailing list
ques...@lists.ntp.org
http://lists.ntp.org/listinfo/questions

Laurent Archambault

unread,
May 12, 2010, 12:20:44 AM5/12/10
to
Hello all,

After many search, i has not find my answer, whay do it mean the status
"XFAC" (?).


I has find this "Extra Fast Attack Crafts", and i am very happy for
this...but in french language it's

very difficult to understand this status "XFAC".
Many thanks - best regards

David J Taylor

unread,
May 12, 2010, 2:09:50 AM5/12/10
to

"Laurent Archambault" <archi....@gmail.com> wrote in message
news:x2w692f4d2f1005112120nb...@mail.gmail.com...

> Hello all,
>
> After many search, i has not find my answer, whay do it mean the status
> "XFAC" (?).
> I has find this "Extra Fast Attack Crafts", and i am very happy for
> this...but in french language it's
> very difficult to understand this status "XFAC".
> Many thanks - best regards

Is this in relation to NTP? One use I found was the string ".XFAC." used
as reference clock source ID (I think).

In other use, you might have scaling for a graph of x-factor and y-factor,
so XFAC and YFAC.

Cheers,
David

Uwe Klein

unread,
May 12, 2010, 4:32:32 AM5/12/10
to
David J Taylor wrote:
>
> "Laurent Archambault" <archi....@gmail.com> wrote in message

>> After many search, i has not find my answer, whay do it mean the status
>> "XFAC" (?).

> Is this in relation to NTP? One use I found was the string ".XFAC."

> used as reference clock source ID (I think).

Looks like, why else would laurent ask here?


>
> In other use, you might have scaling for a graph of x-factor and
> y-factor, so XFAC and YFAC.

Hmpf.

>
> Cheers,
> David

grep -A1 -B2 XFAC $(find . -type f)

./ntp-4.2.6p2-RC3/ntpd/ntp_peer.c- if (peer->dstadr != piface && !(peer->cast_flags &
./ntp-4.2.6p2-RC3/ntpd/ntp_peer.c- MDF_ACAST) && peer->pmode != MODE_BROADCAST)
./ntp-4.2.6p2-RC3/ntpd/ntp_peer.c: peer_clear(peer, "XFAC");
./ntp-4.2.6p2-RC3/ntpd/ntp_peer.c-
--
./ntp-4.2.6p2-RC3/CommitLog- use (peer->pmode == MODE_BROADCAST) rather than
./ntp-4.2.6p2-RC3/CommitLog- (peer->castflags & MDF_BCLNT) to detect ephemeral broadcastclient
./ntp-4.2.6p2-RC3/CommitLog: associations to exempt from peer_clear("XFAC")
./ntp-4.2.6p2-RC3/CommitLog-
--
./ntp-4.2.6p2-RC3/CommitLog- use (peer->pmode == MODE_BROADCAST) rather than
./ntp-4.2.6p2-RC3/CommitLog- (peer->castflags & MDF_BCLNT) to detect broadcastclient
./ntp-4.2.6p2-RC3/CommitLog: associations to exempt from peer_clear("XFAC"), suggested
./ntp-4.2.6p2-RC3/CommitLog- by Frank Kardel.
--
./ntp-4.2.6p2-RC3/CommitLog-
./ntp-4.2.6p2-RC3/CommitLog- ntpd/ntp_p...@1.122, 2009-10-15 03:30:55+00:00,
dave...@shiny.ad.hartbrothers.com +1 -1
./ntp-4.2.6p2-RC3/CommitLog: Do not peer_clear("XFAC") manycast server associations or they will
./ntp-4.2.6p2-RC3/CommitLog- never regain their ACST refid.

now what kind of peer would be tagged "XFAC" ??

inquisitive minds want to know ;-)

uwe

Miroslav Lichvar

unread,
May 12, 2010, 6:33:10 AM5/12/10
to
On 2010-05-12, Uwe Klein <uwe_klein_...@t-online.de> wrote:
>> "Laurent Archambault" <archi....@gmail.com> wrote in message
>
>>> After many search, i has not find my answer, whay do it mean the status
>>> "XFAC" (?).
>
> grep -A1 -B2 XFAC $(find . -type f)
>
> ntp_peer.c- if (peer->dstadr != piface && !(peer->cast_flags &
> ntp_peer.c- MDF_ACAST) && peer->pmode != MODE_BROADCAST)
> ntp_peer.c: peer_clear(peer, "XFAC");
> ntp_peer.c-

> now what kind of peer would be tagged "XFAC" ??
>
> inquisitive minds want to know ;-)

There is a comment before that hunk:

/*
* clear crypto if we change the local address
*/

So it would seem the string is used when the network interface which is
used to send packets to the peer has changed its address and the crypto
stuff had to be reset.

--
Miroslav Lichvar

Dave Hart

unread,
May 12, 2010, 1:44:56 AM5/12/10
to
On Wed, May 12, 2010 at 04:20 UTC, Laurent Archambault wrote:
> After many search, i has not find my answer, whay do it mean the status
> "XFAC" (?).

ntpd resets associations when the local network interface (or, really,
IP address) being used for that peer goes away. This means the clock
filter register of 8 samples is cleared, as are all other timekeeping
stats related to the peer. At that time, the "reference ID" or refid
is changed to XFAC to indicate the cause of the cleared (reset)
association. Typically, that XFAC refid is replaced on receipt of the
next packet from the server or peer.

Since you ask what it means without telling us what problem you are
having, I'm left supposing you are asking because it is not
short-lived in your case. You may be able to work around the problem
by adding -U0 to the ntpd command line so that local network
interfaces (IP addresses) are enumerated by ntpd only once at startup.

Cheers,
Dave Hart

Uwe Klein

unread,
May 12, 2010, 4:18:54 PM5/12/10
to
Hi Dave,

so XFAC stands for
X Inter
F Face
A Association
c Change
??

Does this only happen on interface down/change or at ntp bootup too ?

uwe

Dave Hart

unread,
May 12, 2010, 5:30:50 PM5/12/10
to
On Wed, May 12, 2010 at 20:18 UTC, Uwe Klein wrote:
>
> so XFAC stands for
> X Inter
> F Face
> A Association
> c Change
> ??

I really don't know what was imagined initially. Perhaps eXchange
interFACe? With four capital letters used for these locally-generated
pseudo-refids, there's not a lot of room for clarity.

> Does this only happen on interface down/change or at ntp bootup too ?

Only on interface change. At startup, each association shows .INIT.
as the refid.

Cheers,
Dave Hart

Uwe Klein

unread,
May 13, 2010, 2:05:15 AM5/13/10
to
Thank you.

uwe

0 new messages