Changes to 'work3'
by David Teigland
New branch 'work3' available with the following commits:
commit 85887f4d8e22fbf57e9012ec6dccd2d17a2ff21a
Author: David Teigland <teigland(a)redhat.com>
Date: Mon Oct 1 11:47:46 2012 -0500
sanlock: update man page
to reflect recent changes
Signed-off-by: David Teigland <teigland(a)redhat.com>
commit cc2fd83accfd6c9389dd8adc53ec58ab5b083f82
Author: David Teigland <teigland(a)redhat.com>
Date: Fri Sep 28 14:30:27 2012 -0500
sanlock: change log level for startup and io timeouts
make daemon startup and io timeout messages a warning instead
of an error, so they are written to sanlock.log,
not /var/log/messages (which should be for real errors).
Signed-off-by: David Teigland <teigland(a)redhat.com>
commit 3cc457d215713035cf9aa2024841184913c350d7
Author: Federico Simoncelli <fsimonce(a)redhat.com>
Date: Thu Sep 27 12:06:03 2012 -0400
sanlock: use signal handler for all terminating signals
Signed-off-by: Federico Simoncelli <fsimonce(a)redhat.com>
commit b08e8726c7e854921ec6d8d590c77d4e4d18c4b0
Author: Federico Simoncelli <fsimonce(a)redhat.com>
Date: Thu Sep 27 12:06:02 2012 -0400
python: fix the tuple parsing for py_add_lockspace
Signed-off-by: Federico Simoncelli <fsimonce(a)redhat.com>
commit abaf80bbf21b9c771684def60f73f3f1a287bc20
Author: Federico Simoncelli <fsimonce(a)redhat.com>
Date: Thu Sep 27 15:46:59 2012 -0400
sanlock: configuring the core dump output
All the setrlimit calls have been moved to a new setup_limits function
that is called regardless the presence of the -U and -G options.
The new RLIMIT_CORE limit is set to allow the core dump to be written
in case of a crash (for more details: man core).
Signed-off-by: Federico Simoncelli <fsimonce(a)redhat.com>
commit aff1f7713ca8f3836e1b76611b3222628f9c2b84
Author: David Teigland <teigland(a)redhat.com>
Date: Wed Sep 26 14:02:00 2012 -0500
sanlock: configurable mlockall level
-l 0 -> no mlockall
-l 1 -> mlockall(CURRENT)
-l 2 -> mlockall(CURRENT|FUTURE)
The default is now set to 1 (was 2).
CURRENT|FUTURE results in each pthread_create using
8MB of RSS for the thread stack. CURRENT alone does
not.
Signed-off-by: David Teigland <teigland(a)redhat.com>
11 years, 6 months