Commit graph

149 commits

Author SHA1 Message Date
baude
3611f92ddf BUGFIX: Invalid return codes in kpod
Set the exitsdir for kpod back to /var/run/crio... so kpod can benefit
from the container exit file.

Because 0 is the int32 blank value, kpod needs its own container state
struct with the omitempty removed so it can actually display 0 in
its default json output.

Signed-off-by: baude <bbaude@redhat.com>
2017-10-04 09:34:28 -05:00
Mrunal Patel
7f4f630b98 Merge pull request from runcom/fixies-42
*: bunch of fixes...
2017-09-07 07:37:37 -07:00
Antonio Murdaca
e8553a124d
container_create: better handling of devices
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-09-07 10:13:22 +02:00
Antonio Murdaca
dacc5c3ece
*: correctly wait and close servers
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-09-07 10:13:21 +02:00
Dan Williams
0df30c5319 server: port to github.com/cri-o/ocicni; remove pkg/ocicni
Signed-off-by: Dan Williams <dcbw@redhat.com>
2017-09-06 16:07:50 -05:00
Antonio Murdaca
2ac2832686
server: container_create: store sandbox's ip in annotations
So it can be later retrieved when needed (cadvisor)

Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-09-02 02:05:58 +02:00
Mrunal Patel
49c1fd27ac server: Add infra container store to track them separately
Signed-off-by: Mrunal Patel <mrunalp@gmail.com>
2017-09-01 08:49:08 -07:00
Antonio Murdaca
98da370173
server: set golang runtime max threads
SetMaxThreads from runtime/debug in Golang is called to set max threads
value to 90% of /proc/sys/kernel/threads-max
Should really help performance.

Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-08-23 19:35:08 +02:00
Antonio Murdaca
d56bf090ce
*: update kube vendor to v1.7.4
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-08-22 17:32:14 +02:00
Mrunal Patel
ce17c5214d sandbox: Reduce number of calls to UpdateStatus
Also, we distinguish between container and a pod infra
container in the exit monitor as pod infra containers
aren't stored in the main container index.

Signed-off-by: Mrunal Patel <mrunalp@gmail.com>
2017-08-17 09:01:07 -07:00
Mrunal Patel
30ded83096 Add inotify watcher for container exits
This allows the container list API to return updated status
for exited container without having to call container status first.

Signed-off-by: Mrunal Patel <mpatel@redhat.com>
2017-08-13 08:01:48 -07:00
Mrunal Patel
0e1452a4ec Add a /metrics endpoint for the server
This is for prometheus integration

Signed-off-by: Mrunal Patel <mrunalp@gmail.com>
2017-08-10 13:44:42 -04:00
Daniel J Walsh
63a218a458 Move to new github.com/sirupsen/logrus.
Need to mv to latest released and supported version of logrus
switch github.com/Sirupsen/logrus github.com/sirupsen/logrus

Also vendor in latest containers/storage and containers/image

Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
2017-08-07 11:50:04 -04:00
Nalin Dahyabhai
0b7348b35c Use errors.Cause() when looking at storage errors
The storage library uses github.com/pkg/errors to wrap errors that it
returns from many of its functions, so when passing them to
os.IsNotExist() or comparing them to specific errors defined in the
storage library, unwrap them using errors.Cause().

Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
2017-08-03 15:40:54 -04:00
Ryan Cole
b56da85fc1 Make container server update it's list of containers upon creation
Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-08-01 11:44:57 -04:00
Matthew Heon
9529f565b2 Add option to use file-based locking for libkpod state
Signed-off-by: Matthew Heon <mheon@redhat.com>
2017-07-31 15:58:29 -04:00
Ryan Cole
0c8f106ee8 add basic config struct to libkpod
Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-27 11:12:50 -04:00
Matthew Heon
158893bd1a Move Pod ID and Name indexes into libkpod
Signed-off-by: Matthew Heon <mheon@redhat.com>
2017-07-25 11:46:22 -04:00
Matthew Heon
ece055327a Move Sandbox state into libkpod
Signed-off-by: Matthew Heon <mheon@redhat.com>
2017-07-25 11:46:15 -04:00
Ryan Cole
314d55f284 move shutdown function into libkpod and have server call in to it
Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-24 14:32:02 -04:00
Ryan Cole
4fe17ee16d Update libkpod New()
update libkpod's New() function to use a config struct, and update
server.New() to call into libkpod.New()

Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-24 10:35:39 -04:00
Ryan Cole
49ed4ab710 move reserve/release container name into libkpod
Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-24 10:35:31 -04:00
Ryan Cole
d625e0e468 move containerState to/from disk to libkpod
Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-24 10:35:20 -04:00
Mrunal Patel
7b69d43198 Merge pull request from 14rcole/sandbox-package
Move sandbox into its own package
2017-07-21 07:18:00 -07:00
Ryan Cole
d50dba979e Move sandbox into its own package
If sandbox is in the same package as server, there will be a circular dependency when
kpod create is implemented

Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-21 08:15:55 -04:00
Ryan Cole
680f7a6106 Add kpod push command
Push an image to a specified location, such as to an atomic registry
or a local directory

Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-20 11:12:40 -04:00
Ryan Cole
bd540ac94c Separate container state from sandbox state in server
Move container state data to libkpod, separate from the sandbox
data in server.  However, the move was structured such that sandbox
data could easily be moved over into libkpod in the future

Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-20 08:18:55 -04:00
Ryan Cole
3ddf2efae2 Refactor server sandbox
Make sandbox struct and it methods public, and add accessor methods
for the struct fields

Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-19 08:53:45 -04:00
Ryan Cole
64ad902480 Decouple kubernetes-dependent an non-dependent parts of server
Move non-kubernetes-dependent portions of server struct to libkpod.
So far, only the struct fields have been moved and not their dependent
functions

Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-18 14:23:50 -04:00
Ryan Cole
f3efb850b7 Remove ImageSpec dependency from server
Signed-off-by: Ryan Cole <rcyoalne@gmail.com>
2017-07-17 16:21:56 -04:00
Matthew Heon
4086f2bc37 Replace sync.Mutex with sync.Locker to allow different locks
Signed-off-by: Matthew Heon <mheon@redhat.com>
2017-07-14 16:27:12 -04:00
Mrunal Patel
6b543be50b Rename remnants of ocid to crio
Signed-off-by: Mrunal Patel <mpatel@redhat.com>
2017-07-11 16:19:18 -07:00
Andrew Pilloud
17359e34a6 server: init iptables without kube
Signed-off-by: Andrew Pilloud <andrewpilloud@igneoussystems.com>
2017-06-22 08:51:50 -07:00
Andrew Pilloud
28cd8bde49 server: Hookup kubelet hostport
Signed-off-by: Andrew Pilloud <andrewpilloud@igneoussystems.com>
2017-06-22 08:51:50 -07:00
Samuel Ortiz
0e51bbb778 oci: Support mixing trusted and untrusted workloads
Container runtimes provide different levels of isolation, from kernel
namespaces to hardware virtualization. When starting a specific
container, one may want to decide which level of isolation to use
depending on how much we trust the container workload. Fully verified
and signed containers may not need the hardware isolation layer but e.g.
CI jobs pulling packages from many untrusted sources should probably not
run only on a kernel namespace isolation layer.

Here we allow CRI-O users to define a container runtime for trusted
containers and another one for untrusted containers, and also to define
a general, default trust level. This anticipates future kubelet
implementations that would be able to tag containers as trusted or
untrusted. When missing a kubelet hint, containers are trusted by
default.

A container becomes untrusted if we get a hint in that direction from
kubelet or if the default trust level is set to "untrusted" and the
container is not privileged. In both cases CRI-O will try to use the
untrusted container runtime. For any other cases, it will switch to the
trusted one.

Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
2017-06-15 10:04:36 +02:00
Alexander Larsson
7bb957bf75 Implement non-terminal attach
We use a SOCK_SEQPACKET socket for the attach unix domain socket, which
means the kernel will ensure that the reading side only ever get the
data from one write operation. We use this for frameing, where the
first byte is the pipe that the next bytes are for. We have to make sure
that all reads from the socket are using at least the same size of buffer
as the write side, because otherwise the extra data in the message
will be dropped.

This also adds a stdin pipe for the container, similar to the ones we
use for stdout/err, because we need a way for an attached client
to write to stdin, even if not using a tty.

This fixes https://github.com/kubernetes-incubator/cri-o/issues/569

Signed-off-by: Alexander Larsson <alexl@redhat.com>
2017-06-14 22:59:50 +02:00
Andrew Pilloud
c77b5fbea8 Add stream-address and stream-port flags to crio
Signed-off-by: Andrew Pilloud <andrewpilloud@igneoussystems.com>
2017-06-12 16:12:36 -07:00
Antonio Murdaca
8b53fabcbd
*: support insecure registries
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-06-09 01:04:29 +02:00
Mrunal Patel
065f12490c conmon: Add unix domain socket for attach
Signed-off-by: Mrunal Patel <mpatel@redhat.com>
2017-06-06 07:36:52 -07:00
Mrunal Patel
34357ec7a4 Merge pull request from runcom/streamserver-bindaddress
server: do not use localhost for streaming service
2017-06-03 11:01:43 -07:00
Antonio Murdaca
315c385371
server: do not use localhost for streaming service
The bug is silly if you have a master/node cluster where node is on a
different machine than the master.
The current behavior is to give our addresses like "0.0.0.0:10101". If
you run "kubectl exec ..." from another host, that's not going to work
since on a different host 0.0.0.0 resolves to localhost and kubectl
exec fails with:

error: unable to upgrade connection: 404 page not found

This patch fixes the above by giving our correct addresses for reaching
from outside.

Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-06-03 18:26:33 +02:00
Samuel Ortiz
f15859c79f pkg/annotations: Export CRI-O annotations namespace
Some runtimes like Clear Containers need to interpret the CRI-O
annotations, to distinguish the infra container from the regular one.
Here we export those annotations and use a more standard dotted
namespace for them.

Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
2017-06-01 23:45:44 +02:00
Antonio Murdaca
404194c1fd
server: add nil checks to not panic
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-29 13:16:27 +02:00
Antonio Murdaca
1e9ef65345
server: fix PortForward panic
During "Port forwarding" e2e tests, the following panic happened:

```
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x64981d]

goroutine 52788 [running]:
panic(0x1830ee0, 0xc4200100c0)
        /usr/lib/golang/src/runtime/panic.go:500 +0x1a1
github.com/kubernetes-incubator/cri-o/oci.(*Runtime).UpdateStatus(0xc4202afc00,
0x0, 0x0, 0x0)
        /home/amurdaca/go/src/github.com/kubernetes-incubator/cri-o/oci/oci.go:549
+0x7d
github.com/kubernetes-incubator/cri-o/server.streamService.PortForward(0xc42026e000,
0x0, 0x0, 0x0, 0x0, 0xc420d9af40, 0x40, 0xc400000050, 0x7fe660659a28,
0xc4201cd0e0, ...)
```

The issue is `streamService.PortForward` assumed the first argument to
be the sandbox's infra container ID, thus trying to get it from memory
store using `.state.containers.Get`. Since that ID is of the sandbox
itself, it fails to get the container object from memory and panics in
`UpdateStatus`.

Fix it by looking for the sandbox's infra container ID starting from a
sandbox ID.

Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-28 18:22:46 +02:00
Antonio Murdaca
b4f1cee2a2
server: store and use image's stop signal to stop containers
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-27 10:21:04 +02:00
Antonio Murdaca
3b2e5aa5c2
add a note on server shutdown...
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-18 21:19:52 +02:00
Antonio Murdaca
f00edc2b2d
server: do not remove but stop sandboxes on shutdown
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-18 21:19:52 +02:00
Antonio Murdaca
a41ca975c1
server: restore containers state from disk on startup
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-18 21:19:50 +02:00
Antonio Murdaca
da0b8a6157
server: store containers state on disk
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-18 21:19:50 +02:00
Antonio Murdaca
790c6d891a
server: store creation in containers
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
2017-05-18 18:49:54 +02:00