This test has been flaky in GitHub Actions. This change logs the
values from devmapper to further investigate the issue.
Signed-off-by: Kazuyoshi Kato <katokazu@amazon.com>
btrfs plugin needs CGO support. However on riscv64, cgo
is only support on go1.16 (not released yet).
Instead of setting no_btrfs manually, adding a cgo tag tells
the compiler to skip it automatically.
Signed-off-by: Shengjing Zhu <zhsj@debian.org>
* Currently we rely on making the UVMs sandbox.vhdx in the shim itself instead of this being
made by the snapshotter itself. This change adds a label that affects whether to create the UVMs
scratch layer in the snapshotter itself.
* Adds container scratch size customization. Before adding the computestorage calls
(vendored in with https://github.com/containerd/containerd/pull/4859) there was no way to make a containers
or UVMs scratch size less than the default (20 for containers and 10 for the UVM).
Signed-off-by: Daniel Canter <dcanter@microsoft.com>
Currently we would create a new disk and mount this into the LCOW UVM for every container but there
are certain scenarios where we'd rather just mount a single disk and then have every container share this one
storage space instead of every container having it's own xGB of space to play around with.
This is accomplished by just making a symlink to the disk that we'd like to share and then
using ref counting later on down the stack in hcsshim if we see that we've already mounted this
disk.
Signed-off-by: Daniel Canter <dcanter@microsoft.com>
No need to use the private losetup command line wrapper package.
The generic package provides the same functionality.
Signed-off-by: Peng Tao <bergwolf@hyper.sh>
For LCOW currently we copy (or create) the scratch.vhdx for every single snapshot
so there ends up being a sandbox.vhdx in every directory seemingly unnecessarily. With the default scratch
size of 20GB the size on disk is about 17MB so there's a 17MB overhead per layer plus the time to copy the
file with every snapshot. Only the final sandbox.vhdx is actually used so this would be a nice little
optimization.
For WCOW we essentially do the exact same except copy the blank vhdx from the base layer.
Signed-off-by: Daniel Canter <dcanter@microsoft.com>
When running tests on any modern distro, this assumption will work. If
we need to make it work with kernels where we don't append this option
it will require some more involved changes.
Signed-off-by: Phil Estes <estesp@linux.vnet.ibm.com>
Previously there wwasn't a way to pass any labels to snapshotters as the wrapper
around WithNewSnapshot didn't have a parm to pass them in.
Signed-off-by: Daniel Canter <dcanter@microsoft.com>
Put the overlay plugin in a separate package to allow the overlay package to be
used without needing to import and initialize the plugin.
Signed-off-by: Derek McGowan <derek@mcg.dev>
This allows configuring the location of the overlayfs snapshotter by
adding the following in config.toml
```
[plugins]
[plugins.overlayfs]
root_path = "/custom_location"
```
This is useful to isolate disk i/o for overlayfs from the rest of
containerd and prevent containers saturating disk i/o from negatively
affecting containerd operations and cause timeouts.
Signed-off-by: Ashray Jain <ashrayj@palantir.com>
The rollback mechanism is implemented by calling deleteDevice() and
RemoveDevice(). But RemoveDevice() is internally calling
deleteDevice() as well.
Since a device will be deleted by first deleteDevice(),
RemoveDevice() always will see ENODATA. The specific error must be
ignored to remove the device's metadata correctly.
Signed-off-by: Kazuyoshi Kato <katokazu@amazon.com>
kernel version > 4.13rc1 support index=on feature, it will be failed
with EBUSY when trying to mount.
Related: https://github.com/moby/moby/pull/37993
Signed-off-by: Rudy Zhang <rudyflyzhang@gmail.com>
Dependencies may be switching to use the new `%w` formatting
option to wrap errors; switching to use `errors.Is()` makes
sure that we are still able to unwrap the error and detect the
underlying cause.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The issue beblow happens several times beforing the root
cause found:
1. A `fdisk -l` process has being hung up for a long time;
2. A image layer snapshot device is visiable to dmsetup, which
should *not* happen because it should be deactivated after
`Commit()`;
The backtrace of `fdisk` is always the same over time:
```bash
[<ffffffff810bbc6a>] io_schedule+0x2a/0x80
[<ffffffff81295a3f>] do_blockdev_direct_IO+0x1e9f/0x2f10
[<ffffffff81296aea>] __blockdev_direct_IO+0x3a/0x40
[<ffffffff81290e43>] blkdev_direct_IO+0x43/0x50
[<ffffffff811b8a14>] generic_file_read_iter+0x374/0x960
[<ffffffff81291ad5>] blkdev_read_iter+0x35/0x40
[<ffffffff8125229b>] new_sync_read+0xfb/0x240
[<ffffffff81252406>] __vfs_read+0x26/0x40
[<ffffffff81252b96>] vfs_read+0x96/0x130
[<ffffffff812540e5>] SyS_read+0x55/0xc0
[<ffffffff81003c04>] do_syscall_64+0x74/0x180
```
The root cause is, in Commit(), there's a race window between
`SuspendDevice()` and `DeactivateDevice()`, which may cause the
IOs of a process or command like `fdisk` on the "suspended" device
hang up forever. It has twofold:
1. The IOs suspends on the devices;
2. The device is in `Suspended` state, because it's deactivated with
`deferred` flag and without `force` flag;
So they cannot make progress.
One reproducer is:
1. enlarge the race window by putting sleep seconds there;
2. run `while true; do sudo fdisk -l; sleep 0.5; done` on one terminal;
3. and pull image on another terminal;
Fixes it by:
1. Resume the devices again after flushing IO by suspend;
2. Remove device without `deferred` flag;
Fix: #4234
Signed-off-by: Eric Ren <renzhen@linux.alibaba.com>
For remote snapshotter cases it's quite often there is need to pass extra info
from client (for instance - registry URL to query remote layer from, credentials, etc).
This commit slightly extends WithPullSnapshotter to pass extra labels to a snapshotter.
Signed-off-by: Maksym Pavlenko <makpav@amazon.com>
Snapshots GC takes use of pruneBranch() function to remove snapshots,
but GC will stop if snapshotter.Remove() returns error and the error
number is not ErrFailedPrecondition. This results in thousands of
dm snapshots not deleted if one snapshot is not deleted, due to
errors like "contains a filesystem in use".
So return ErrFailedPrecondition error number in Remove() function where
appropriate, and let GC process go on collecting other snapshots.
Fix: #3923
Signed-off-by: Eryu Guan <eguan@linux.alibaba.com>
Signed-off-by: Eric Ren <renzhen.rz@linux.alibaba.com>
Cleanup is an optional method a snapshotter may implement.
Cleanup can be used to cleanup resources after a snapshot
has been removed. This function allows a snapshotter to defer
longer resource cleanup until after snapshot removals are
completed. Adding this to the API allows proxy snapshotters
to leverage this enhancement.
Signed-off-by: Derek McGowan <derek@mcgstyle.net>
