On Tue, Jul 29, 2014 at 11:07:37AM -0700, Daniel Jung wrote:
Hi,
Running centos 6 with sssd-1.8.0-32.el6.x86_64 and sssd-client-1.8.0-32.el6.x86_64.
These are very old versions, first recommendation is to upgrade. In 6.5, we have a heavily patched 1.9 branch (pretty much equivalent to 1.9.6 upstream).
This particular problem seems to pop now and then. I looked thru the maillist/googled and there seems to be number of posts with similar symptoms, but didnt see exact bug number or recommended actions to resolve this issue. Could you advise please? The problem resolved after restarting sssd.
sssd.log (Tue Jul 29 16:46:06 2014) [sssd] [mt_svc_sigkill] (0x0010): [LDAP][22259] is not responding to SIGTERM. Sending SIGKILL.
I suspect you're running with enumerate=true ? If so, then I suspect your backend took too long to save all entries and was killed by the monitor. Either disable enumeration (preferred) or increase the 'timeout' parameter a bit.