Right now we seem to have 3 locks.
- devinfo.lock
This is a per device lock
- metaData.devicesLock
This is supposedely protecting map of devices.
- Global DeviceSet lock
This is protecting map of devices as well as serializing calls to libdevmapper.
Semantics of per devices lock and global deviceset lock seem to be very clear.
Even ordering between these two locks has been defined properly.
What is not clear is the need and ordering of metaData.devicesLock. Looks like
this lock is not necessary and global DeviceSet lock should be used to
protect map of devices as it is part of DeviceSet.
This patchset gets rid of metaData.devicesLock and instead uses DeviceSet
lock to protect map of devices.
Also at couple of places during initialization takes devices.Lock(). That
is not strictly necessary as there is supposed to be one thread of execution
during initializaiton. Still it makes the code clearer.
I think this makes code more clear and easier to understand and easier to
make further changes.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
maxDeviceID is upper limit on device Id thin pool can support. Right now
we have this check only during startup. It is a good idea to move this
check in loadMetadata so that any time a device file is loaded and if it
is corrupted and device Id is more than maxDevieceID, it will be detected
right then and there.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Use deactivateDevice() instead of removeDevice() directly. This will make
sure for device deletion, deferred removal is used if user has configured
it in. Also this makes reading code litle easier as there is single function
to remove a device and that is deactivateDevice().
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
If a device is still mounted at the time of DeleteDevice(), that means
higher layers have not called Put() properly on the device and are trying
to delete it. This is a bug in the code where Get() and Put() have not been
properly paired up. Fail device deletion if it is still mounted.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Exists() and HasDevice() just check if device file exists or not. It does
not say anything about if device is mounted or not. Fix comments.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
device has map (device.Devices), contains valid devices and we skip all
the files which are not device files. transaction metadata file is not
device file. Skip this file when devices files are being read and loaded
into map.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Implement basic interfaces to write custom routers that can be plugged
to the server. Remove server coupling with the daemon.
Signed-off-by: David Calavera <david.calavera@gmail.com>
This passes through the container hostname to HCS, which in Windows Server
2016 TP4 will set the container's hostname in the registry before starting
it. This will be silently ignored by TP3.
Signed-off-by: John Starks <jostarks@microsoft.com>
Although having a request ID available throughout the codebase is very
valuable, the impact of requiring a Context as an argument to every
function in the codepath of an API request, is too significant and was
not properly understood at the time of the review.
Furthermore, mixing API-layer code with non-API-layer code makes the
latter usable only by API-layer code (one that has a notion of Context).
This reverts commit de41640435, reversing
changes made to 7daeecd42d.
Signed-off-by: Tibor Vass <tibor@docker.com>
Conflicts:
api/server/container.go
builder/internals.go
daemon/container_unix.go
daemon/create.go
This reverts commit ff92f45be4, reversing
changes made to 80e31df3b6.
Reverting to make the next revert easier.
Signed-off-by: Tibor Vass <tibor@docker.com>
This fixes the case where directory is removed in
aufs and then the same layer is imported to a
different graphdriver.
Currently when you do `rm -rf /foo && mkdir /foo`
in a layer in aufs the files under `foo` would
only be be hidden on aufs.
The problems with this fix:
1) When a new diff is recreated from non-aufs driver
the `opq` files would not be there. This should not
mean layer differences for the user but still
different content in the tar (one would have one
`opq` file, the others would have `.wh.*` for every
file inside that folder). This difference also only
happens if the tar-split file isn’t stored for the
layer.
2) New files that have the filenames before `.wh..wh..opq`
when they are sorted do not get picked up by non-aufs
graphdrivers. Fixing this would require a bigger
refactoring that is planned in the future.
Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
Fixes an issue where `VOLUME some_name:/foo` would be parsed as a named
volume, allowing access from the builder to any volume on the host.
This makes sure that named volumes must always be passed in as a bind.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Before this patch libcontainer badly errored out with `invalid
argument` or `numerical result out of range` while trying to write
to cpuset.cpus or cpuset.mems with an invalid value provided.
This patch adds validation to --cpuset-cpus and --cpuset-mems flag along with
validation based on system's available cpus/mems before starting a container.
Signed-off-by: Antonio Murdaca <runcom@linux.com>
Use `pkg/discovery` to provide nodes discovery between daemon instances.
The functionality is driven by two different command-line flags: the
experimental `--cluster-store` (previously `--kv-store`) and
`--cluster-advertise`. It can be used in two ways by interested
components:
1. Externally by calling the `/info` API and examining the cluster store
field. The `pkg/discovery` package can then be used to hit the same
endpoint and watch for appearing or disappearing nodes. That is the
method that will for example be used by Swarm.
2. Internally by using the `Daemon.discoveryWatcher` instance. That is
the method that will for example be used by libnetwork.
Signed-off-by: Arnaud Porterie <arnaud.porterie@docker.com>
* Thanks to the Default gateway service in libnetwork, we dont have to add
containers explicitly to secondary public network. This is handled
automatically regardless of the primary network driver.
* Fixed the URL convention for kv-store to be aligned with the upcoming
changes to discovery URL
* Also, in order to bring consistency between external and internal network
drivers, we moved the driver configs via controller Init.
Signed-off-by: Madhu Venugopal <madhu@docker.com>
Avoid creating a global context object that will be used while the daemon is running.
Not only this object won't ever be garbage collected, but it won't ever be used for anything else than creating other contexts in each request. I think it's a bad practive to have something like this sprawling aroud the code.
This change removes that global object and initializes a context in the cases we don't have already one, like shutting down the server.
This also removes a bunch of context arguments from functions that did nothing with it.
Signed-off-by: David Calavera <david.calavera@gmail.com>
The mount syscall does not handle string flags like "noatime",
we must use bitmasks like MS_NOATIME instead.
pkg/mount.Mount already handles this work.
Signed-off-by: Carl Henrik Lunde <chlunde@ping.uio.no>
This PR adds a "request ID" to each event generated, the 'docker events'
stream now looks like this:
```
2015-09-10T15:02:50.000000000-07:00 [reqid: c01e3534ddca] de7c5d4ca927253cf4e978ee9c4545161e406e9b5a14617efb52c658b249174a: (from ubuntu) create
```
Note the `[reqID: c01e3534ddca]` part, that's new.
Each HTTP request will generate its own unique ID. So, if you do a
`docker build` you'll see a series of events all with the same reqID.
This allow for log processing tools to determine which events are all related
to the same http request.
I didn't propigate the context to all possible funcs in the daemon,
I decided to just do the ones that needed it in order to get the reqID
into the events. I'd like to have people review this direction first, and
if we're ok with it then I'll make sure we're consistent about when
we pass around the context - IOW, make sure that all funcs at the same level
have a context passed in even if they don't call the log funcs - this will
ensure we're consistent w/o passing it around for all calls unnecessarily.
ping @icecrime @calavera @crosbymichael
Signed-off-by: Doug Davis <dug@us.ibm.com>
- Print the mount table as in /proc/self/mountinfo
- Do not exit prematurely when one of the ipc mounts doesn't exist.
- Do not exit prematurely when one of the ipc mounts cannot be unmounted.
- Add a unit test to see if the cleanup really works.
- Use syscall.MNT_DETACH to cleanup mounts after a crash.
- Unmount IPC mounts when the daemon unregisters an old running container.
Signed-off-by: David Calavera <david.calavera@gmail.com>
- Add unit tests to make sure the functionality is correct.
- Add FilterByDriver to allow filtering volumes by driver, for future
`volume ls` filtering and whatnot.
Signed-off-by: David Calavera <david.calavera@gmail.com>
If an invalid logger address is provided on daemon start it will
silently fail. As syslog driver is doing, this check should be done on
daemon start and prevent it from starting even in other drivers.
This patch also adds integration tests for this behavior.
Signed-off-by: Antonio Murdaca <runcom@linux.com>
`docker rename foo ''` would result in:
```
usage: docker rename OLD_NAME NEW_NAME
```
which is the old engine's way of return errors - yes that's in the
daemon code. So I fixed that error msg to just be normal.
While doing that I noticed that using an empty string for the
source container name failed but didn't print any error message at all.
This is because we would generate a URL like: ../containers//rename/..
which would cause a 301 redirect to ../containers/rename/..
however the CLI code doesn't actually deal with 301's - it just ignores
them and returns back to the CLI code/caller.
Rather than changing the CLI to deal with 3xx error codes, which would
probably be a good thing to do in a follow-on PR, for this immediate
issue I just added a cli-side check for empty strings for both old and
new names. This way we catch it even before we hit the daemon.
API callers will get a 404, assuming they follow the 301, for the
case of the src being empty, and the new error msg when the destination
is empty - so we should be good now.
Add tests for both cases too.
Signed-off-by: Doug Davis <dug@us.ibm.com>
This way provide both Time and TimeNano in the event. For the display of
the JSONMessage, use either, but prefer TimeNano Proving only TimeNano
would break Subscribers that are using the `Time` field, so both are set
for backwards compatibility.
The events logging uses nano formatting, but only provides a Unix()
time, therefor ordering may get lost in the output. Example:
```
2015-09-15T14:18:51.000000000-04:00 ee46febd64ac629f7de9cd8bf58582e6f263d97ff46896adc5b508db804682da: (from busybox) resize
2015-09-15T14:18:51.000000000-04:00 a78c9149b1c0474502a117efaa814541926c2ae6ec3c76607e1c931b84c3a44b: (from busybox) resize
```
By having a field just for Nano time, when set, the marshalling back to
`time.Unix(sec int64, nsec int64)` has zeros exactly where it needs to.
This does not break any existing use of jsonmessage.JSONMessage, but now
allows for use of `UnixNano()` and get event formatting that has
distinguishable order. Example:
```
2015-09-15T15:37:23.810295632-04:00 6adcf8ed9f5f5ec059a915466cd1cde86a18b4a085fc3af405e9cc9fecbbbbaf: (from busybox) resize
2015-09-15T15:37:23.810412202-04:00 6b7c5bfdc3f902096f5a91e628f21bd4b56e32590c5b4b97044aafc005ddcb0d: (from busybox) resize
```
Including tests for TimeNano and updated event API reference doc.
Signed-off-by: Vincent Batts <vbatts@redhat.com>
GET /containers/json route used to reply with and empty array `[]` when no
containers where available. Daemon containers list refactor introduced
this bug by declaring an empty slice istead of initializing it as well
and it was now replying with `null`.
Signed-off-by: Antonio Murdaca <runcom@linux.com>
- refactor to make it easier to split the api in the future
- addition to check the existing test case and make sure it contains
some expected output
Signed-off-by: Morgan Bauer <mbauer@us.ibm.com>
Using @mavenugo's patch for enabling the libcontainer pre-start hook to
be used for network namespace initialization (correcting the conflict
with user namespaces); updated the boolean check to the more generic
SupportsHooks() name, and fixed the hook state function signature.
Signed-off-by: Madhu Venugopal <madhu@docker.com>
Docker-DCO-1.1-Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com> (github: estesp)
With go1.5's concurrency, the use of a goroutine in Log'ing events was
causing the resulting events to not be in order.
Signed-off-by: Vincent Batts <vbatts@redhat.com>
For now docker stats will sum the rxbytes, txbytes, etc. of all
the interfaces.
It is OK for the output of CLI `docker stats` but not good for
the API response, especially when the container is in sereval
subnets.
It's better to leave these origianl data to user.
Signed-off-by: Hu Keping <hukeping@huawei.com>
Jess Frazelle
Jessica Frazelle
Lei Jitang
Darren Stahl
Vivek Goyal
David Calavera
Tibor Vass
Antonio Murdaca
Alexander Morozov
John Starks
Tibor Vass
Tonis Tiigi
Brian Goff
Chun Chen
John Howard
Antonio Murdaca
Arnaud Porterie
Madhu Venugopal
Morgan Bauer
Michael Crosby
Carl Henrik Lunde
Doug Davis
qhuang
Alexandre Beslic
Hu Keping
Evan Hazlett
Zhang Kun
Vincent Batts
Shijiang Wei
Philip Monroe
Tim Hockin
Sebastiaan van Stijn
Alexander Morozov