Date: Fri, 2 Aug 2002 17:16:03 +0200 (CEST) From: Martin Karsten <Martin.Karsten@KOM.tu-darmstadt.de> To: freebsd-net@FreeBSD.ORG Subject: device polling breaks ntp? Message-ID: <200208021516.g72FG4M18743@KOM.tu-darmstadt.de>
next in thread | raw e-mail | index | archive | help
I have reasons to believe that network device polling breaks the functionality of ntpd. I'm using FreeBSD 4.5 and ntp-4.1.71. When device polling is enabled *before* ntpd is started (particularly, when device polling is enabled at boot time), ntpd seems to be unable to properly synchronize with a reference clock. While I understand that the increased packet delay variance introduced by network device polling might affect ntpd's algorithms to estimate the network delay, it also seems to break clock synchronization with a locally attached GPS-receiver. When network device polling is enabled *after* ntpd is started, it does not cause immediate problems and I have not further investigated what exactly happens in this case. In any case, disabling sysctl kern.polling.enable and restarting ntpd results in proper operation of ntpd. Is this behaviour known and is there a workaround? Thanks, Martin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200208021516.g72FG4M18743>