Tema
05.03.2010, 18:15
Никак не получается правильно настроить NTP на CS1000M MG Rel.6.0. Дока перечитана несколько раз и все сделано по ней, но не работает :(
CS 1000 system level primary and, optionally, secondary NTP servers
are configured on Linux system elements that are part of this CS1000
system. The secondary NTP server would act as a backup for the
primary NTP server, and normally synchronize time with the primary
NTP server and then try with other external sources. The default is
that the element on which Element Manager is running is set as the
CS 1000 primary NTP server, but that can be altered. All other Linux
system elements (including EM if applicable) will synchronize to these
CS 1000 NTP servers. Configuration is done by EM and pushed to all
Linux elements.
The CS 1000 primary and secondary NTP servers can source their
time in two ways:
— The CS 1000 primary and secondary NTP servers use their internal
hardware clocks. The date/time has to be set using Base Manager
on the primary (assuming that the secondary NTP server will sync
time from the primary NTP server in normal operation).
— External NTP clock sources are used. The internal system primary
and secondary NTP servers are synchronized from external clock
sources, up to 10, with optional single key security. The secondary
NTP server would normally synchronize with the primary NTP
server, and only synchronized with the external sources if the
primary is not available.
и чуть ниже:
NTP clients running on Linux base elements which are "Not a clock
server", as well as on the VxWorks-based Call Server, get their time
source from the Primary internal NTP server, or from the Secondary
internal NTP server, if the Primary does not respond to polling by the other
NTP clients in the CS 1000 system.
То есть получается, что linux'овый сигнальник (он у меня один) является Primary internal NTP server, берет время с внешнего NTP-сервера (который находится в подсети ELAN), а Call Server берет время с этого сигнального сервера.
И вот что получается на самом деле: станция сихронизирует время от сигнальника (2 раза в сутки), но сигнальник не может синхронизировать время от внешнего сервера времени. В логах сигнальника появляется вот такая ошибка:
38 Mar 4 14:30:52 linuxbase: (INFO) Base: basesettime(61): PID[5396]: New time is not applied because the server is either NTP Client or NTP Server.
39 Mar 5 02:30:43 linuxbase: (INFO) Base: basesettime(61): PID[29799]: New time is not applied because the server is either NTP Client or NTP Server.
40 Mar 5 14:30:35 linuxbase: (INFO) Base: basesettime(61): PID[22547]: New time is not applied because the server is either NTP Client or NTP Server.
Что я делаю неправильно?
CS 1000 system level primary and, optionally, secondary NTP servers
are configured on Linux system elements that are part of this CS1000
system. The secondary NTP server would act as a backup for the
primary NTP server, and normally synchronize time with the primary
NTP server and then try with other external sources. The default is
that the element on which Element Manager is running is set as the
CS 1000 primary NTP server, but that can be altered. All other Linux
system elements (including EM if applicable) will synchronize to these
CS 1000 NTP servers. Configuration is done by EM and pushed to all
Linux elements.
The CS 1000 primary and secondary NTP servers can source their
time in two ways:
— The CS 1000 primary and secondary NTP servers use their internal
hardware clocks. The date/time has to be set using Base Manager
on the primary (assuming that the secondary NTP server will sync
time from the primary NTP server in normal operation).
— External NTP clock sources are used. The internal system primary
and secondary NTP servers are synchronized from external clock
sources, up to 10, with optional single key security. The secondary
NTP server would normally synchronize with the primary NTP
server, and only synchronized with the external sources if the
primary is not available.
и чуть ниже:
NTP clients running on Linux base elements which are "Not a clock
server", as well as on the VxWorks-based Call Server, get their time
source from the Primary internal NTP server, or from the Secondary
internal NTP server, if the Primary does not respond to polling by the other
NTP clients in the CS 1000 system.
То есть получается, что linux'овый сигнальник (он у меня один) является Primary internal NTP server, берет время с внешнего NTP-сервера (который находится в подсети ELAN), а Call Server берет время с этого сигнального сервера.
И вот что получается на самом деле: станция сихронизирует время от сигнальника (2 раза в сутки), но сигнальник не может синхронизировать время от внешнего сервера времени. В логах сигнальника появляется вот такая ошибка:
38 Mar 4 14:30:52 linuxbase: (INFO) Base: basesettime(61): PID[5396]: New time is not applied because the server is either NTP Client or NTP Server.
39 Mar 5 02:30:43 linuxbase: (INFO) Base: basesettime(61): PID[29799]: New time is not applied because the server is either NTP Client or NTP Server.
40 Mar 5 14:30:35 linuxbase: (INFO) Base: basesettime(61): PID[22547]: New time is not applied because the server is either NTP Client or NTP Server.
Что я делаю неправильно?