[PATCH master] Document master daemon's shutdown behaviour

4 views
Skip to first unread message

Michael Hanselmann

unread,
Nov 14, 2012, 3:03:23 AM11/14/12
to ganeti...@googlegroups.com
This wasn't clearly documented until now.
---
man/ganeti-masterd.rst | 12 ++++++++++++
1 files changed, 12 insertions(+), 0 deletions(-)

diff --git a/man/ganeti-masterd.rst b/man/ganeti-masterd.rst
index 5159277..2502d65 100644
--- a/man/ganeti-masterd.rst
+++ b/man/ganeti-masterd.rst
@@ -55,6 +55,18 @@ skip this process. The option requires interactive confirmation, as
having two masters on the same cluster is a very dangerous
situation and will most likely lead to data loss.

+SHUTDOWN
+~~~~~~~~
+
+When the master daemon receives a ``SIGINT`` (Ctrl-C) or ``SIGTERM``
+signal, it stops accepting new jobs and prepares to shut down shortly
+after jobs running at the time finish. Meanwhile it still responds to
+LUXI requests. Pending jobs are re-queued after the daemon is started
+again. When a second signal is received, the master daemon terminates
+without waiting (it might still need to wait for some threads to
+finish). Note that the latter, as well as sending ``SIGKILL``, may leave
+the cluster in an inconsistent state.
+
JOB QUEUE
~~~~~~~~~

--
1.7.7.3

Iustin Pop

unread,
Nov 14, 2012, 3:19:31 AM11/14/12
to Michael Hanselmann, ganeti...@googlegroups.com
LGTM, but this specific behaviour is not good (race conditions between
two processes trying to tell masterd to shutdown). I'll file a bug for
correct it - we had the same behaviour in hbal and people complained
about it.

thanks,
iustin
Reply all
Reply to author
Forward
0 new messages