]>
git.proxmox.com Git - mirror_lxcfs.git/log
Serge Hallyn [Mon, 16 Nov 2015 20:18:26 +0000 (14:18 -0600)]
Merge pull request #53 from hallyn/reboot
Add a hook to make reboot sleep 0.5s
Serge Hallyn [Thu, 12 Nov 2015 17:15:06 +0000 (11:15 -0600)]
Add a hook to make reboot sleep 0.5s (v2)
To ensure that the new container does not get cached directory entry
information from before we removed the container's cgroup using cgroupfs.
Changelog: make sure lxc.reboot.hook shows up in 'make dist'
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 12 Nov 2015 07:41:52 +0000 (01:41 -0600)]
Limit caching to 0.5s
If a cgroup is deleted or chmoded using the underlying cgroupfs, then we
want to minimize the amount of time during which we get stale info. At the
same time, we don't want to do away with caching in the fuse kernel module
altogether, since calling out to userspace is expensive.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Wed, 11 Nov 2015 01:38:40 +0000 (19:38 -0600)]
chown cgroup files after chowning dirs
cgmanager used to do this for us.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Wed, 11 Nov 2015 00:42:34 +0000 (18:42 -0600)]
mkdir: don't do it as setresid
First of all the code was wrong anyway.
Secondly, doing mkdir under setres{ug}id won't work because that
won't respect privilege that callers in user namespaces may have.
We check the permissions meticulously anyway, so do the mkdir as
root and then chown the new cgroup.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Teemu Grönqvist [Sat, 31 Oct 2015 19:56:07 +0000 (21:56 +0200)]
Implement SwapTotal and SwapFree support for /proc/meminfo
[ Serge: s/cgm_/cgfs_/ ]
Signed-off-by: Teemu Grönqvist <teemu.gronqvist@net9.fi>
Serge Hallyn [Mon, 9 Nov 2015 20:48:10 +0000 (14:48 -0600)]
Makefile.am: add cgfs.h to lxcfs_SOURCES
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 9 Nov 2015 05:26:33 +0000 (23:26 -0600)]
Use cpuacct info for /proc/uptime idle field
As future work, the calculation of the reaper pid (to find its cpuacct cgroup)
and its reaper age could be combined into one step.
Closes #50
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 9 Nov 2015 05:13:50 +0000 (23:13 -0600)]
Use find_mounted_controller in get_pid_cgroup
So that get_pid_cgroup("cpu", pid) works when cpu is comounted.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 9 Nov 2015 05:12:30 +0000 (23:12 -0600)]
find_mounted_controller: support comounted controllers
So that when an internal fn asks for a cgroup value in 'cpu',
it finds it under 'cpu,cpuacct'.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 2 Nov 2015 22:56:36 +0000 (16:56 -0600)]
convert to using cgfs instead of cgmanager
remove lxcfs_mkdir - it is not needed since we don't need a clean dbus connection
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 04:33:16 +0000 (23:33 -0500)]
drop support for precise's glib
since we can't run in travis anyway
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 04:32:25 +0000 (23:32 -0500)]
ok, don't run the tests after all bc we'd need to modprobe fuse
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 04:27:12 +0000 (23:27 -0500)]
travis: satisfy older glib by calling g_type_init
Sad, really.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 04:21:49 +0000 (23:21 -0500)]
travis: forget it for now
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 02:42:33 +0000 (21:42 -0500)]
travis: no really
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 02:37:28 +0000 (21:37 -0500)]
travis: noninteractive upgrade
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 02:21:45 +0000 (21:21 -0500)]
travis: do a dist-upgrade...?
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 02:15:45 +0000 (21:15 -0500)]
final travis fixes (i hope)
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 02:15:26 +0000 (21:15 -0500)]
use LXCFSDIR in meminfo hierarchy test
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 01:57:53 +0000 (20:57 -0500)]
add more travis (trusty) requirements
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 01:52:44 +0000 (20:52 -0500)]
travis: install coreutils
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 31 Oct 2015 01:26:42 +0000 (20:26 -0500)]
run tests as root
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 30 Oct 2015 23:30:56 +0000 (18:30 -0500)]
/proc/meminfo: show the lowest limit amongst our ancestors
If we are in /a/b/c, and b is limited to 500k, then c's limit_in_bytes
will not reflect the 500k, although that will be enforced for us. So
check our lineage for the lowest limit.
Closes #44
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 30 Oct 2015 18:03:34 +0000 (13:03 -0500)]
actually run tests from travis
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 30 Oct 2015 17:58:12 +0000 (12:58 -0500)]
support running tests from travis
Add a main.sh which can run a self-confined testsuite.
Note that without having lxcfs_mkdir installed, we cannot mkdir as
non-root user.
Update the test scripts to support using a lxcfs mountpoint other
than /var/lib/lxcfs
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 30 Oct 2015 17:57:50 +0000 (12:57 -0500)]
enable travis
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 30 Oct 2015 17:29:18 +0000 (12:29 -0500)]
don't let idletime be > reaperage
This is not a good way to do this. We should decide on a proper
heuristic. We could take something like reaperage * (idletime/total_uptime),
but that doesn't scale for how much our own container used the cpu nor
for time.
I will open a github issue to fix this.
However as it currently stands the test_proc testcase was failing, this
at least lets it pass.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Stéphane Graber [Wed, 28 Oct 2015 22:32:54 +0000 (18:32 -0400)]
Merge pull request #49 from hallyn/offset1
fix corner cases in uptime and diskstat read
Serge Hallyn [Wed, 28 Oct 2015 20:41:45 +0000 (20:41 +0000)]
fix corner cases in uptime and diskstat read
Closes #33
The code for these (a shortcut version of the other proc_*_read ones) was
doing snprintf(buf, size, ...). If the user only requests one byte, we
just served them a trailing 0. Fix that.
We weren't handling reads with offset in these cases either. Fix that.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Wed, 28 Oct 2015 18:36:18 +0000 (13:36 -0500)]
Merge pull request #48 from bmiklautz/uptime_format
Change uptime format
Serge Hallyn [Tue, 27 Oct 2015 20:49:05 +0000 (15:49 -0500)]
update .gitignore
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Tue, 27 Oct 2015 20:48:46 +0000 (15:48 -0500)]
move lxcfs_mkdir under libexecdir
Note, this will require us to add
dh_auto_configure -- --libdir=\$${prefix}/lib/$(DEB_HOST_MULTIARCH) \
--libexecdir=\$${prefix}/lib/$(DEB_HOST_MULTIARCH)
to debian/rules.
Closes #46
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Bernhard Miklautz [Wed, 28 Oct 2015 13:07:28 +0000 (14:07 +0100)]
Change "uptime" format
/proc/uptime has the format %lu.%02lu %lu.%02lu. The format used by
lxcfs doesn't consider the nano second portion of the uptime which might
cause programs that rely on that format to fail.
This commit adapts the uptime format to match the kernels by adding
trailing dummy values (.0) to the uptime and idle time values.
The parsing of /proc/uptime was updated.
Signed-off-by: Bernhard Miklautz <bernhard.miklautz@shacknet.at>
Serge Hallyn [Tue, 27 Oct 2015 20:43:30 +0000 (15:43 -0500)]
configure.ac: drop libcgmanager dependency
Closes #45
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 26 Oct 2015 20:59:25 +0000 (15:59 -0500)]
configure.ac: v0.11
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 26 Oct 2015 19:37:31 +0000 (14:37 -0500)]
Remove a memory leak from getattr
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 22 Oct 2015 23:30:57 +0000 (18:30 -0500)]
remove stale comments, update some
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 22 Oct 2015 23:01:07 +0000 (18:01 -0500)]
fix memory leak
each key also has an allocated char*, so use list_key to free the key.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Wolfgang Bumiller [Wed, 21 Oct 2015 09:37:08 +0000 (11:37 +0200)]
lxc mount hook: whitespace fixup
Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Wolfgang Bumiller [Wed, 21 Oct 2015 09:37:07 +0000 (11:37 +0200)]
lxc mount hook: add missing space before ]
Otherwise [ interprets the ] as part of the path and also
errors about a missing ']'
Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 16 Oct 2015 19:44:29 +0000 (14:44 -0500)]
swap out libnih and libnih-dbus for glib
The motivation is to make threading possible, to hopefully greatly
speed up systemd startup inside containers.
This required converting all the nih-ified memory tracking. (Some
of this can probably be done smarter in a more glib-friendly way, i.e.
avoiding doing a glib string alloc followed by strdup followed by
freeing the glib string)
We open a single dbus connection for all threads to use. If that
connection is closed (i.e. cgmanager exits / restarts) the first
task to find it so takes a mutex and attempts to reconnect, once
per second, until it is reconnected.
When creating a directory for non-root user, execute a new binary
to get a clean dbus session as that user.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 15 Oct 2015 13:29:56 +0000 (08:29 -0500)]
allow systemd to escape its init.scope cgroup
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Tue, 13 Oct 2015 20:34:41 +0000 (15:34 -0500)]
Merge pull request #38 from sn-x/master
Fixed issue #37
Aljaz Meznaric [Fri, 9 Oct 2015 09:54:12 +0000 (11:54 +0200)]
Fixed issue #37
Stéphane Graber [Mon, 14 Sep 2015 14:56:02 +0000 (10:56 -0400)]
Merge pull request #34 from ralt/patch-1
Better explain what lxcfs does
Florian Margaine [Mon, 14 Sep 2015 10:22:29 +0000 (12:22 +0200)]
Better explain what lxcfs does
At first glance, I thought this project was a custom fs for the hosts, not for the containers themselves.
Hopefully this line will clear up the misunderstanding for future
readers, and show that this project is *really* useful!
Signed-off-by: Florian Margaine <florian@margaine.com>
Stéphane Graber [Thu, 3 Sep 2015 22:55:54 +0000 (18:55 -0400)]
configure.ac: v0.10
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Serge Hallyn [Thu, 3 Sep 2015 13:46:52 +0000 (08:46 -0500)]
fix symlink
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 3 Sep 2015 04:00:11 +0000 (23:00 -0500)]
lxc mount hook: implement stgraber's suggestions
revert to using sh and use symlinks instead of extra mounts.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 3 Sep 2015 03:28:59 +0000 (22:28 -0500)]
Update lxc mount hook to ovver a mount per co-mounted controller
If cpu,cpuacct are comounted, then in addition to cpu,cpuacct make
mounts for 'cpu' and 'cpuacct'.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Wed, 24 Jun 2015 23:26:06 +0000 (18:26 -0500)]
lxc mount hook: don't mount lxcfs fs for any controller already setup
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Wed, 24 Jun 2015 23:15:34 +0000 (18:15 -0500)]
don't thread
The pieces are there, but for now it's still not safe with libdbus
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 5 Jun 2015 04:18:20 +0000 (23:18 -0500)]
Return host's meminfo file if no memory cgroup
If memory cgroup is not available (that is, we can't find a
memory cgroup for the reading task), then just return the
contents of the host's /proc/meminfo. (Same for all proc file
reads)
Closes #30
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Wed, 20 May 2015 15:56:46 +0000 (08:56 -0700)]
fix two threading issues
Make sure to prep dbus for threading before we start.
And use _exit() any time we are exiting from a forked child. This is
to avoid calling the at_exit() functions. Once a thread in the main
program has called nih_error_init(), this registers an at_exit fn which
asserts that the nih_context not be null - but after we fork, if libnih
is built with --enable-threading, then the nih_context is in fact null.
The only way to clear the atexit fns would be to exec(). So call
_exit() instead of exit(), because _exit() avoids calling the atexit
fns.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Tue, 19 May 2015 21:27:05 +0000 (14:27 -0700)]
use threads when safe
libnih, when not built with --enable-threading, cannot be safely
used by a threaded application. Detect whether it is built to be
threadsafe using a new libnih helper, and, if so, run threaded by
(a) not passing '-s' to fuse, and (b) making the dbus connection
and detected api version thread-local.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sat, 9 May 2015 00:52:28 +0000 (19:52 -0500)]
configure.ac: v0.9
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Stéphane Graber [Fri, 8 May 2015 21:09:38 +0000 (21:09 +0000)]
Merge pull request #26 from mikemccracken/maybe-fix-24
Add error handling for snprintf calls inside *_read() funcs
Michael McCracken [Fri, 8 May 2015 20:24:03 +0000 (16:24 -0400)]
inside *_read() funcs, return 0 for snprintf errors and truncations
Signed-off-by: Michael McCracken <mike.mccracken@canonical.com>
Michael McCracken [Fri, 8 May 2015 08:46:02 +0000 (04:46 -0400)]
check return value from snprintf to avoid FUSE errors. Potentially fixes #24.
Signed-off-by: Michael McCracken <mike.mccracken@canonical.com>
Serge Hallyn [Thu, 7 May 2015 19:39:34 +0000 (14:39 -0500)]
configure.ac: release 0.8
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 7 May 2015 16:25:52 +0000 (11:25 -0500)]
implement --version
Because this is needed for help2man to determine the program name.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Thu, 30 Apr 2015 11:07:34 +0000 (13:07 +0200)]
Fix the check for cpu being in cpuset range
Also add a testcase to make sure that the new function is correct.
In order to share the cpuset range checking code with with the
test, move it into cpuset.c. Not sure whether we want that in a
utils.c instead.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 20 Apr 2015 13:40:39 +0000 (08:40 -0500)]
Add some more sanity checks
Limit string size in sscanf, convert some sprintfs to snprintfs
(to make them future-proof), and free d in case of error.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sun, 19 Apr 2015 04:51:03 +0000 (23:51 -0500)]
free d at program end
to silence warnings from memory checkers
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Sun, 19 Apr 2015 04:03:45 +0000 (23:03 -0500)]
Make sure that that '/cgroup' and the controller are sep'd by /
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 17 Apr 2015 21:13:33 +0000 (16:13 -0500)]
cg_read: add ending newline
Closes #23
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 17 Apr 2015 20:39:35 +0000 (15:39 -0500)]
use direct_io
This will prevent a file read for over getattr.
File sizes show up as 0, but they do so in cgroupfs and procfs as well.
Closes #20.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 17 Apr 2015 20:28:04 +0000 (15:28 -0500)]
Provide the fuse options we want ourselves
continue to allow the options which in the past we basically required
callers to provide ("-s -f -o allow_other"), exit on anything else.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 10 Apr 2015 18:54:46 +0000 (13:54 -0500)]
Cleanup whitespace
and remove commented out debug statement
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 10 Apr 2015 18:52:02 +0000 (13:52 -0500)]
Merge pull request #22 from hustcat/cache.1
cache for proc file
Ye Yin [Fri, 10 Apr 2015 07:19:44 +0000 (15:19 +0800)]
info as parent when malloc memory for file buf
reduce one blank for /proc/stat
Signed-off-by: Ye Yin <eyniy@qq.com>
Serge Hallyn [Thu, 9 Apr 2015 21:18:06 +0000 (16:18 -0500)]
Merge pull request #19 from hallyn/cache.1
Cache.1
Ye Yin [Thu, 9 Apr 2015 11:01:10 +0000 (19:01 +0800)]
add cache for proc file, for support multiple read
Signed-off-by: Ye Yin <eyniy@qq.com>
Stéphane Graber [Tue, 7 Apr 2015 20:17:40 +0000 (16:17 -0400)]
LXC hook fixes
- Detects whether cgroups are enabled in the container's config
- Don't mount a second tmpfs on /sys/fs/cgroup
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Serge Hallyn [Tue, 7 Apr 2015 03:10:17 +0000 (22:10 -0500)]
cache: use file_dir object as parent for alloc'ed objects
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 6 Apr 2015 22:24:11 +0000 (17:24 -0500)]
don't check for cgroup perms when opendir'ing /cgroup
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 6 Apr 2015 22:11:39 +0000 (17:11 -0500)]
add a file_info cache pointer for procfile reads
This doesn't actually use them yet, but now fi->fh->buf can be
used in https://github.com/lxc/lxcfs/pull/18
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 6 Apr 2015 22:03:14 +0000 (17:03 -0500)]
add cache types for each procfile type that we provide
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 6 Apr 2015 21:35:20 +0000 (16:35 -0500)]
check that file and dir entries don't get mixed up
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 6 Apr 2015 21:30:54 +0000 (16:30 -0500)]
cache verified controller/cgroup info at cg file read/write
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 6 Apr 2015 21:13:54 +0000 (16:13 -0500)]
cache opendir work and re-use it at readdir
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 3 Apr 2015 20:09:09 +0000 (15:09 -0500)]
release 0.7
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 3 Apr 2015 18:25:57 +0000 (13:25 -0500)]
be robust to glibc fork failures
kill child and retry when __glibc_fork() child hangs on an assert
(newpid != ppid) failure.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 3 Apr 2015 01:44:39 +0000 (20:44 -0500)]
timeout the reception of a scm_cred
Else if the peer is killed at a strange time we will hang forever waiting.
Also don't kill children - always just let them time out
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Fri, 3 Apr 2015 01:28:58 +0000 (20:28 -0500)]
nih_local's must be initialized as NULL
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Wed, 18 Mar 2015 14:59:02 +0000 (09:59 -0500)]
Merge pull request #16 from hustcat/diskstats
Add /proc/diskstats
Ye Yin [Wed, 11 Mar 2015 08:33:08 +0000 (16:33 +0800)]
Add /proc/diskstats
Signed-off-by: Ye Yin <eyniy@qq.com>
Serge Hallyn [Wed, 25 Feb 2015 05:00:27 +0000 (23:00 -0600)]
Merge pull request #15 from steelbrain/master
Fix formatting of README
steelbrain [Wed, 25 Feb 2015 04:49:57 +0000 (23:49 -0500)]
Fix formatting of README
Signed-off-by: Anees Iqbal <aneesiqbalbhatti@gmail.com>
Stéphane Graber [Tue, 17 Feb 2015 04:42:33 +0000 (23:42 -0500)]
Fix test_proc following the stat fix
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Serge Hallyn [Tue, 17 Feb 2015 03:56:10 +0000 (21:56 -0600)]
Release LXCFS 0.6
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Christian Brauner [Tue, 17 Feb 2015 01:41:54 +0000 (19:41 -0600)]
Show cpu-average in /proc/stat and start cup numbering at 0
Signed-off-by: Christian Brauner <christianvanbrauner@gmail.com>
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Serge Hallyn [Mon, 16 Feb 2015 22:33:53 +0000 (16:33 -0600)]
clean up after several functions
which weren't freeing line of closing an open file.
Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Stéphane Graber [Sat, 31 Jan 2015 12:47:45 +0000 (13:47 +0100)]
Update README
Close #10
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Stéphane Graber [Wed, 28 Jan 2015 14:45:54 +0000 (15:45 +0100)]
Release LXCFS 0.5
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Stéphane Graber [Wed, 28 Jan 2015 14:35:06 +0000 (15:35 +0100)]
hook: Don't spam /etc/mtab
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Stéphane Graber [Wed, 28 Jan 2015 09:07:05 +0000 (10:07 +0100)]
Make the LXC hook more robust
Don't fail if /sys/fs/cgroup or some proc files don't exist.
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Stéphane Graber [Tue, 27 Jan 2015 05:37:53 +0000 (05:37 +0000)]
Update gitignore
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>
Stéphane Graber [Mon, 26 Jan 2015 22:07:21 +0000 (17:07 -0500)]
Better detect the cgmanager version
Signed-off-by: Stéphane Graber <stgraber@ubuntu.com>