(Fri May 23 00:30:33 2014) [sssd] [service_send_ping] (4): Pinging LDAP (Fri May 23 00:30:33 2014) [sssd] [service_send_ping] (4): Pinging nss (Fri May 23 00:30:33 2014) [sssd] [service_send_ping] (4): Pinging pam (Fri May 23 00:30:33 2014) [sssd] [ping_check] (4): Service LDAP replied to ping (Fri May 23 00:30:33 2014) [sssd] [ping_check] (4): Service pam replied to ping (Fri May 23 00:30:33 2014) [sssd] [ping_check] (4): Service nss replied to ping
I see that based on timeout setting, there is 10 secs timeout setting. What exactly does Pinging LDAP do? because i dont see any packets coming thru the line using tcpdump from the hosts listed in the URI.
Thanks
On (22/05/14 15:32), Daniel Jung wrote:
(Fri May 23 00:30:33 2014) [sssd] [service_send_ping] (4): Pinging LDAP (Fri May 23 00:30:33 2014) [sssd] [service_send_ping] (4): Pinging nss (Fri May 23 00:30:33 2014) [sssd] [service_send_ping] (4): Pinging pam (Fri May 23 00:30:33 2014) [sssd] [ping_check] (4): Service LDAP replied to ping (Fri May 23 00:30:33 2014) [sssd] [ping_check] (4): Service pam replied to ping (Fri May 23 00:30:33 2014) [sssd] [ping_check] (4): Service nss replied to ping
I see that based on timeout setting, there is 10 secs timeout setting. What exactly does Pinging LDAP do? because i dont see any packets coming thru the line using tcpdump from the hosts listed in the URI.
SSSD is not single process application. There are various processes which servers different purposes. The main process (sssd) monitor all child processes and try to restart unresponsive process. The main process does not use ICMP ping. For all communication, unix sockets are used.
man page says: If a service is not responding to ping checks (see the “timeout” option), it is first sent the SIGTERM signal that instructs it to quit gracefully. If the service does not terminate after “force_timeout” seconds, the monitor will forcibly shut it down by sending a SIGKILL signal.
LS
sssd-users@lists.fedorahosted.org