> In SC06 there is "GPS A & B disagree" and "GPS B is inactive". How can > they disagree when only one is working?
Ah my apologies, I made a mistake with the table. SC06 has both GPS sources active but their measurements disagree. This is an edge case but it has happened in practice. > Is the server trying to poll itself? What is that supposed to do? I had this in there hoping that it would help Server A resolve its local refid in orphan mode. It doesn't seem to have helped. It was my understanding that chrony will ignore servers that resolve to the localhost. I will remove those lines if they are useless. > Which scenarios break when you don't use any of the trust options? Without the trust options in place, I'm unable to meet SC06 since the clients won't select a server, Server B will select GPS B as its source. > That indicates chronyd doesn't know the local IP address for the orhan > source. Is it running on normal Linux? Can you check strace if > received messages have the IP_PKTINFO cmsg? The server computers are raspberry pi 4s running raspbian buster. Chrony version info: ``` $ chronyd -v chronyd (chrony) version 3.4 (+CMDMON +NTP +REFCLOCK +RTC +PRIVDROP +SCFILTER +SIGND +ASYNCDNS +SECHASH +IPV6 -DEBUG) ``` I am seeing some syscalls to `recvmmsg` with the IP_PKTINFO cmsg type. On Wed, Jan 22, 2025 at 3:43 AM Miroslav Lichvar <mlich...@redhat.com> wrote: > On Tue, Jan 21, 2025 at 04:28:00PM -0500, William Bergen wrote: > > I’ve been spending quite a bit of time trying to configure the system to > > meet my needs but I can’t seem to satisfy every scenario simultaneously. > I > > know this is complicated by the fact that there are only 2 servers in the > > network. I’ve tried to compile my requirements into a table of scenarios > > and desired outcomes. > > [image: scenarios.png] > > In SC06 there is "GPS A & B disagree" and "GPS B is inactive". How can > they disagree when only one is working? > > > So far, my biggest trouble has been meeting SC04 AND SC06 with the same > > configuration. This is where I’m at right now with the relevant sections > of > > the configuration: > > Server A > > > > rtcsyncallow 192.168.5 > > local stratum 10 orphan > > makestep 1 -1 > > refclock SHM 0 refid GPS delay 0.2 trust > > server 192.168.5.1 > > Is the server trying to poll itself? What is that supposed to do? > > Which scenarios break when you don't use any of the trust options? > > > This config setup is the closest I have gotten to achieving my desired > > outcomes. I thought that the orphan option for the local refclock was > > correct in this scenario but I keep receiving an error when testing > > SC05: Unknown > > local refid in orphan mode. > > That indicates chronyd doesn't know the local IP address for the orhan > source. Is it running on normal Linux? Can you check strace if > received messages have the IP_PKTINFO cmsg? > > -- > Miroslav Lichvar > > > -- > To unsubscribe email chrony-users-requ...@chrony.tuxfamily.org > with "unsubscribe" in the subject. > For help email chrony-users-requ...@chrony.tuxfamily.org > with "help" in the subject. > Trouble? Email listmas...@chrony.tuxfamily.org. > >
