The ability to save and verify base device UUID (#13896) introduced a
situation where the initialization would panic when removing the device
returns EBUSY.
Functions `verifyBaseDeviceUUID` and `saveBaseDeviceUUID` now take the
lock on the `DeviceSet`, which solves the problem as `removeDevice`
assumes it owns the lock.
Signed-off-by: Arnaud Porterie <arnaud.porterie@docker.com>
Often it happens that docker is not able to shutdown/remove the thin
pool it created because some device has leaked into some mount name
space. That means device is in use and that means pool can't be removed.
Docker will leave pool as it is and exit. Later when user starts the
docker, it finds pool is already there and docker uses it. But docker
does not know it is same pool which is using the loop devices. Now
docker thinks loop devices are not being used. That means it does not
display the data correctly in "docker info", giving user wrong information.
This patch tries to detect if loop devices as created by docker are
being used for pool and fills in the right details in "docker info".
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
DeviceMapper must be explicitly selected because the Docker binary might not be linked to the right devmapper library.
With this change, Docker fails fast if the driver detection finds the devicemapper directory but the driver is not the default option.
The option `override_udev_sync_check` doesn't make sense anymore, since the user must be explicit to select devicemapper, so it's being removed.
Docker fails to use devicemapper only if Docker has been built statically unless the option was explicit.
Signed-off-by: David Calavera <david.calavera@gmail.com>
Export metadata for container and image in docker-inspect when overlay
graphdriver is in use. Right now it is done only for devicemapper graph
driver.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
It is easy for one to use docker for a while, shut it down and restart
docker with different set of storage options for device mapper driver
which will effectively change the thin pool. That means any of the
metadata stored in /var/lib/docker/devicemapper/metadata/ is not valid
for the new pool and user will run into various kind of issues like
container not found in the pool etc.
Users think that their images or containers are lost but it might just
be the case of configuration issue. People might use wrong metadata
with wrong pool.
To detect such situations, save UUID of base image and once docker
starts later, query and compare the UUID of base image with the
stored one. If they don't match, fail the initialization with the
error that UUID failed to match.
That way user will be forced to cleanup /var/lib/docker/ directory
and start docker again.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Export image/container metadata stored in graph driver. Right now 3 fields
DeviceId, DeviceSize and DeviceName are being exported from devicemapper.
Other graph drivers can export fields as they see fit.
This data can be used to mount the thin device outside of docker and tools
can look into image/container and do some kind of inspection.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Previously the cache was only updated once on startup, because the graph
code only check for filesystems on startup. However this breaks the API as it
was supposed and so unit tests.
Fixes#13142
Signed-off-by: Jörg Thalheim <joerg@higgsboson.tk>
The docker graph call driver.Exists() on initialisation for each filesystem in
the graph. This results will results in a lot `zfs get all` commands. To reduce
this, retrieve all descend filesystem at startup and cache it for later checks
Signed-off-by: Jörg Thalheim <joerg@higgsboson.tk>
instead of let zfs automaticly mount datasets, mount them on demand using mount(2).
This speed up this graph driver in 2 ways:
- less zfs processes needed to start a container
- /proc/mounts get smaller, so zfs userspace tools has less to read (which can
a significant amount of data as the number of layer grows)
This ways it can be also ensured that the correct mountpoint is always used.
Signed-off-by: Jörg Thalheim <joerg@higgsboson.tk>
Right now devicemapper mounts thin device using online discards by default
and passes mount option "discard". Generally people discourage usage of
online discards as they can be a drain on performance. Instead it is
recommended to use fstrim once in a while to reclaim the space.
In case of containers, we recommend to keep data volumes separate. So
there might not be lot of rm, unlink operations going on and there might
not be lot of space being freed by containers. So it might not matter
much if we don't reclaim that free space in pool.
User can still pass mount option explicitly using dm.mountopt=discard to
enable discards if they would like to.
So this is more like setting the containers by default for better performance
instead of better space efficiency in pool. And user can change the behavior
if they don't like default behavior.
Reported-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
If device is being reactivated before it could go away and deferred
deactivation is scheduled on it, cancel it.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
This will help with debugging as one could just do "docker info" and figure
out of deferred removal is enabled or not.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Provide a new command line knob dm.deferred_device_removal which will enable
deferred device deactivation if driver and library support it.
This patch also checks for library support and driver version.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Before this, a storage driver would be defaulted to based on the
priority list, and only print a warning if there is state from other
drivers.
This meant a reordering of priority list would "break" users in an
upgrade of docker, such that there images in the prior driver's state
were now invisible.
With this change, prior state is scanned, and if present that driver is
preferred.
As such, we can reorder the priority list, and after an upgrade,
existing installs with prior drivers can have a contiguous experience,
while fresh installs may default to a driver in the new priority list.
Ref: https://github.com/docker/docker/pull/11962#issuecomment-88274858
Signed-off-by: Vincent Batts <vbatts@redhat.com>
Signed-off-by: Megan Kostick <mkostick@us.ibm.com>
Alphabetize FSMagic list to make more human-readable.
Signed-off-by: Megan Kostick <mkostick@us.ibm.com>
This provides an override for forcing the daemon to still attempt
running the devicemapper driver even when udev sync is not supported.
Intended to be a very clear impairment for those choosing to use it. If
udev sync is false, there will still be an error in the daemon logs,
even when the override is in place. The docs have an explicit WARNING.
Including link to the docs for users that encounter this daemon error
during an upgrade.
Signed-off-by: Vincent Batts <vbatts@redhat.com>
Right now we try device removal at the interval of 10ms and keep on trying
till either device is removed or 10 seconds are over. That means if device
is busy, we will try 1000 times in those 10 seconds.
Sounds too high a frequency of deivce removal retrial. All the logs are
filled easily. I think it is a good idea to slow down a bit and retry at
the interval of 100ms instead of 10ms.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
During device removal, we are first waiting for device to close() in a tight
loop for 10 seconds. I am not sure why do we need it. First of all we come
here once the umount() is successful so device should be free. For some reason
of device is temporarily busy, then removeDevice() logic retries device removal
logic in a loop for 10 seconds and that should cover it. Can't see why one
more 10 seoncds loop is required before attempting device removal.
One loop should be able to cover all the temporary device busy conditions and
if condition is not temporary then 10 seconds loop is not going to help anyway.
So instead of two loops of 10 seconds each, I am converting it to a single
loop of 20 seconds. May be 10 second loop is good enough but for now I am
keeping it 20 seconds to avoid any regressions.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Currently in device removal path (device deactivation), we wait
for 10 seconds for devive to actually go away. waitRemove().
In current code this is not required. If dm removal task has completed
and one has done the wait on udev cookie, then device is gone and there
is no need to write another loop to wait for device removal.
This patch removes the waitRemove() which waits for 10 seconds after
device removal. This seems unnecessary.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
devmapper graph driver retries device removal 1000 times in case of failure
and if this fills up console with 1000 messages (when daemon is running in
debug mode). So remove these debug messages.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
There are issues with libdm logging. Right now if docker daemon is run
in debug mode, logging by libdm is too verbose. And if a device can't
be removed, thousands of messages fill the console and one can not see
what's going on.
This patch removes devicemapper.LogInitVerbose() call as that call will
only work if docker was not registering its own log handler with libdm.
For some reason docker registers one with libdm and libdm hands over
all the messages to docker (including debug ones). And now it is up to
devmapper backend to figure out which ones should go to console and
which ones should not.
So by default log only fatal messages from libdm. One can easily modify
the code to change it for debugging purposes.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
It's about time to let folks not hit 'vfs', when 'overlay' is supported
on their kernel. Especially now that v3.18.y is a long-term kernel.
Signed-off-by: Vincent Batts <vbatts@redhat.com>
Automatically detect support for aufs `dirperm1` option and apply it.
`dirperm1` tells aufs to check the permission bits of the directory on the
topmost branch and ignore the permission bits on all lower branches.
It can be used to fix aufs' permission bug (i.e., upper layer having
broader mask than the lower layer).
More information about the bug can be found at https://github.com/docker/docker/issues/783
`dirperm1` man page is at: http://aufs.sourceforge.net/aufs3/man.html
Signed-off-by: Daniel, Dao Quang Minh <dqminh89@gmail.com>
We removed it, because upstream removed it. But now it will be coming
back, so work with it either way.
Signed-off-by: Vincent Batts <vbatts@redhat.com>
They say we should only use the BTRFS_LIB_VERSION
They will no longer support this since it had to be managed manually
Docker-DCO-1.1-Signed-off-by: Dan Walsh <dwalsh@redhat.com> (github: rhatdan)
In several cases graphdriver were just returning the low-level syscall
error and that was making it all the way up to the daemon logs and in
many cases was difficult to tell it was even coming from the graphdriver
at all.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
daemon/volumes.go
This SetFileCon call made no sense, it was changing the labels of any
directory mounted into the containers SELinux label. If it came from me,
then I apologize since it is a huge bug.
The Volumes Mount code should optionally do this, but it should not always
happen, and should never happen on a --privileged container.
The change to
daemon/graphdriver/vfs/driver.go, is a simplification since this it not
a relabel, it is only a setting of the shared label for docker volumes.
Docker-DCO-1.1-Signed-off-by: Dan Walsh <dwalsh@redhat.com> (github: rhatdan)
Vincent Batts
John Howard
Arnaud Porterie
Jessie Frazelle
David Calavera
Vivek Goyal
Mary Anthony
Brian Goff
Jörg Thalheim
Tibor Vass
mapk0y
Alexey Guskov
Alexander Morozov
Michael Crosby
Lei Jitang
Alexander Morozov
Burke Libbey
unclejack
Antonio Murdaca
Megan Kostick
Vincent Batts
Michael Crosby
Peter Choi
Daniel, Dao Quang Minh
Anes Hasicic
Jimmy Puckett
Dan Walsh
Michal Minar
Deng Guangxing
Qiang Huang
Martijn Dwars
Phil Estes
Mabin
Srini Brahmaroutu
Peter Volpe
Tibor Vass
Sven Dowideit
Jessica Frazelle