This change no longer requires the use of a UtlityVM on Windows to convert the
layer tar to an ext4 vhd for LCOW. This has a significant performance boost
that makes linux/amd64 layer extraction comparable to native Linux performance.
Signed-off-by: Justin Terry (VM) <juterry@microsoft.com>
The default duration of cpu profile collection in net/http/pprof is 30
seconds. User should have chance to set the specific duration for the
collection.
Signed-off-by: Wei Fu <fuweid89@gmail.com>
On Windows because of the way the log pipe is forwarded to the shim there is a
condition where the pipe listener may not yet be active when a client tries to
connect. To handle this case we allow polling on the file and rety on pipe not
found. This limits the pipe not found retry to 5 seconds but leaves the connect
timeout alone as if there is a listener we want to connect to it normally.
Signed-off-by: Justin Terry (VM) <juterry@microsoft.com>
containerd should cache empty label for docker schema1 image.
if not, the original empty layer will be non-empty layer and the image
config will be changed too. in this case, the image ID will be changed.
check the blob empty label to avoid changing image ID when repull docker
schema1 image.
Signed-off-by: Wei Fu <fuweid89@gmail.com>
The OCI image specification includes a `StopSignal` field in the image
configuration, denoting the system call signal to be sent to the
container to exit. This commit adds a new `WithImageStopSignal` container
option that can be used for storing the `StopSignal` field as a label on
the container. This commit also adjusts `ctr run` to call
`WithImageStopSignal` and `ctr tasks kill` to send the signal stored in
that label by default.
Signed-off-by: Samuel Karp <skarp@amazon.com>
Revendors to Microsoft/hcsshim v0.7.5 that added support for logging all
runhcs.exe commands via Windows named pipes. This now launches all runhcs.exe
commands and forwards debug logging to the containerd-shim-runhcs log when
with --debug.
Signed-off-by: Justin Terry (VM) <juterry@microsoft.com>
Signed-off-by: Lifubang <lifubang@aliyun.com>
fix some errors
use typeurl.UnmarshalAny instead json.Unmarshal to interface{}
Signed-off-by: Lifubang <lifubang@aliyun.com>
Signed-off-by: Michael Crosby <crosbymichael@gmail.com>
A panic was seen related to the buffer being reset in
one goroutine while being read in another. In the case
of pigz an early cancellation will cause the reader to
close, resetting the buffer and signaling the process
to shut down, but races since the process must stop
reading before the reset otherwise the a panic may occur.
This fix guarantees that the bufio is always reset and
returned to the pool on the same goroutine that is
doing the read. If a buffer is not fully read the
buffered reader should just be discarded and not
returned back to the pool.
Signed-off-by: Derek McGowan <derek@mcgstyle.net>
Avoid filtering manifests by platform when a non-index
manifest is provided to the Manifest function.
Signed-off-by: Derek McGowan <derek@mcgstyle.net>
