When a network id is used to create a container we translate it to use the
name internally for the db. The network aliases are also stored with the
network name as key so we have to also translate them for the db.
Also removed some outdated skips from the e2e tests.
Fixes#11285
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
To match Docker's behavior, in the `--net=host` case, we need to
use the host's `/etc/hosts` file, unmodified (without adding an
entry for the container). We will still respect hosts from
`--add-host` but will not make any automatic changes.
Fortuntely, this is strictly a matter of removal and refactoring
as we already base our `/etc/hosts` on the host's version - just
need to remove the code that added entries when net=host was set.
Fixes#10319
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
e2e test failures are rife with messages like:
Expected 1 to equal 0
These make me cry. They're anti-helpful, requiring the reader
to dive into the source code to figure out what those numbers
mean.
Solution: Go tests have a '.Should(Exit(NNN))' mechanism. I
don't know if it spits out a better diagnostic (I have no way
to run e2e tests on my laptop), but I have to fantasize that
it will, and given the state of our flakes I assume that at
least one test will fail and give me the opportunity to see
what the error message looks like.
THIS IS NOT REVIEWABLE CODE. There is no way for a human
to review it. Don't bother. Maybe look at a few random
ones for sanity. If you want to really review, here is
a reproducer of what I did:
cd test/e2e
! positive assertions. The second is the same as the first,
! with the addition of (unnecessary) parentheses because
! some invocations were written that way. The third is BeZero().
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.To\(Equal\((\d+)\)\)/Expect($1).Should(Exit($2))/' *_test.go
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.To\(\(Equal\((\d+)\)\)\)/Expect($1).Should(Exit($2))/' *_test.go
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.To\(BeZero\(\)\)/Expect($1).Should(Exit(0))/' *_test.go
! Same as above, but handles three non-numeric exit codes
! in run_exit_test.go
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.To\(Equal\((\S+)\)\)/Expect($1).Should(Exit($2))/' *_test.go
! negative assertions. Difference is the spelling of 'To(Not)',
! 'ToNot', and 'NotTo'. I assume those are all the same.
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.To\(Not\(Equal\((0)\)\)\)/Expect($1).To(ExitWithError())/' *_test.go
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.ToNot\(Equal\((0)\)\)/Expect($1).To(ExitWithError())/' *_test.go
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.NotTo\(Equal\((0)\)\)/Expect($1).To(ExitWithError())/' *_test.go
! negative, old use of BeZero()
perl -pi -e 's/Expect\((\S+)\.ExitCode\(\)\)\.ToNot\(BeZero\(\)\)/Expect($1).Should(ExitWithError())/' *_test.go
Run those on a clean copy of main branch (at the same branch
point as my PR, of course), then diff against a checked-out
copy of my PR. There should be no differences. Then all you
have to review is that my replacements above are sane.
UPDATE: nope, that's not enough, you also need to add gomega/gexec
to the files that don't have it:
perl -pi -e '$_ .= "$1/gexec\"\n" if m!^(.*/onsi/gomega)"!' $(grep -L gomega/gexec $(git log -1 --stat | awk '$1 ~ /test\/e2e\// { print $1}'))
UPDATE 2: hand-edit run_volume_test.go
UPDATE 3: sigh, add WaitWithDefaultTimeout() to a couple of places
UPDATE 4: skip a test due to bug #10935 (race condition)
Signed-off-by: Ed Santiago <santiago@redhat.com>
When the containers.conf field "NetNS" is set to "Bridge" and the
"RootlessNetworking" field is set to "cni", Podman will now
handle rootless in the same way it does root - all containers
will be joined to a default CNI network, instead of exclusively
using slirp4netns.
If no CNI default network config is present for the user, one
will be auto-generated (this also works for root, but it won't be
nearly as common there since the package should already ship a
config).
I eventually hope to remove the "NetNS=Bridge" bit from
containers.conf, but let's get something in for Brent to work
with.
Signed-off-by: Matthew Heon <mheon@redhat.com>
The initial version of libimage changed the order of layers which has
now been restored to remain backwards compatible.
Further changes:
* Fix a bug in the journald logging which requires to strip trailing
new lines from the message. The system tests did not pass due to
empty new lines. Triggered by changing the default logger to
journald in containers/common.
* Fix another bug in the journald logging which embedded the container
ID inside the message rather than the specifid field. That surfaced
in a preceeding whitespace of each log line which broke the system
tests.
* Alter the system tests to make sure that the k8s-file and the
journald logging drivers are executed.
* A number of e2e tests have been changed to force the k8s-file driver
to make them pass when running inside a root container.
* Increase the timeout in a kill test which seems to take longer now.
Reasons are unknown. Tests passed earlier and no signal-related
changes happend. It may be CI VM flake since some system tests but
other flaked.
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
We missed bumping the go module, so let's do it now :)
* Automated go code with github.com/sirkon/go-imports-rename
* Manually via `vgrep podman/v2` the rest
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
The mtu default value is currently forced to 65520.
This let the user control it using the config key network_cmd_options,
i.e.: network_cmd_options=["mtu=9000"]
Signed-off-by: bitstrings <pino.silvaggio@gmail.com>
Make sure we pass the ip and mac address as CNI_ARGS to
the cnitool which is executed in the rootless-cni-infra
container.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
The existing logic (Range > 0) always triggered, because range is
guaranteed to be at least 1 (a single port has a range of 1, a
two port range (e.g. 80-81) has a range of 2, and so on). As such
this could cause ports that had a host port assigned to them by
the user to randomly assign one instead.
Fixes#8650Fixes#8651
Signed-off-by: Matthew Heon <mheon@redhat.com>
Docker defines an option of "default" which means to
use the default network. We should support this with
the same code path as --network="".
This is important for compatibility with the Docker API.
Fixes: https://github.com/containers/podman/issues/8544
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
As described in issue #8507 this commit contains a breaking
change which is not wanted in v2.2.
We can discuss later if we want this in 3.0 or not.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
We allow a container to be connected to several cni networks
but only if they are listed comma sperated. This is not intuitive
for users especially since the flag parsing allows multiple string
flags but only would take the last value. see: spf13/pflag#72
Also get rid of the extra parsing logic for pods. The invalid options
are already handled by `pkg/specgen`.
A test is added to prevent a future regression.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
The --hostname and containername should always be added to containers.
Added some tests to make sure you can always ping the hostname and container
name from within the container.
Fixes: https://github.com/containers/podman/issues/8095
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
ed identified that the dnsname integration test does not use a unique
name and therefore cannot be cleaned up. this was made worse by a
improper defer statement to remove the network should the test fail.
Signed-off-by: baude <bbaude@redhat.com>
When creating a container in a pod the podname was always set as
the dns entry. This is incorrect when the container is not part
of the pods network namespace. This happend both rootful and
rootless. To fix this check if we are part of the pods network
namespace and if not use the container name as dns entry.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
Detached containers and detach keys are only created with the podman run, i
exec, and start commands. We do not store the detach key sequence or the
detach flags in the database, nor does Docker. The current code was ignoreing
these fields but documenting that they can be used.
Fix podman create man page and --help output to no longer indicate that
--detach and --detach-keys works.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
This does not match Docker, which does not add hostname in this
case, but it seems harmless enough.
Fixes#8095
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
When a container uses --net=host the default hostname is set to
the host's hostname. However, we were not creating any entries
in `/etc/hosts` despite having a hostname, which is incorrect.
This hostname, for Docker compat, will always be the hostname of
the host system, not the container, and will be assigned to IP
127.0.1.1 (not the standard localhost address).
Also, when `--hostname` and `--net=host` are both passed, still
use the hostname from `--hostname`, not the host's hostname (we
still use the host's hostname by default in this case if the
`--hostname` flag is not passed).
Fixes#8054
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
In Podman 1.9.3, `podman run -p 80` would assign port 80 in the
container to a random port on the host. In Podman 2.0 and up, it
assigned Port 80 in the container to Port 80 on the host. This is
an easy fix, fortunately - just need to remove the bit that
assumed host port, if not given, should be set to container port.
We also had a test for the bad behavior, so fix it to test for
the correct way of doing things.
Fixes#7947
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
We do not populate the hostname field with the IP Address
when running within a user namespace.
Fixes https://github.com/containers/podman/issues/7490
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Remove ones that are not needed.
Document those that should be there.
Document those that should be fixed.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Problem: if either of the two "podman network create" tests
fail, all subsequent retries will also fail because the
created network has not been cleaned up (so "network create"
will fail with EEXIST).
Solution: run "podman network rm" as deferred cleanup instead
of in each test.
This is NOT a fix for #7583 - it is just a way to allow
ginkgo to retry a failing test.
Signed-off-by: Ed Santiago <santiago@redhat.com>
Fixes: 4c75fe3f70 ("fix pod creation with "new:" syntax")
Commit 4c75fe3f70 passes all net options to the pod but forgot
to unset the options for the container creation. This leads to
erros when using flags like `--ip` since we tried setting
the ip on the pod and container which obviously fails.
I didn't notice the bug because we don't throw an error when
specifing port bindings on a container which joins the pods
network namespace. (#7373)
Also allow the use of `--hostname` and pass that option to the
pod and unset it for the container. The container has to use
the pods hostname anyway. This would error otherwise.
Added tests to prevent regression.
Signed-off-by: Paul Holzinger <paul.holzinger@web.de>
This adds support for the --cidr parameter that is supported
by slirp4netns since v0.3.0. This allows the user to change
the ip range that is used for the network inside the container.
Signed-off-by: Adis Hamzić <adis@hamzadis.com>
Docker and CNI have very different ideas of what 0.0.0.0 means.
Docker takes it to be 0.0.0.0/0 - that is, bind to every IPv4
address on the host. CNI (and, thus, root Podman) take it to mean
the literal IP 0.0.0.0. Instead, CNI interprets the empty string
("") as "bind to all IPs".
We could ask CNI to change, but given this is established
behavior, that's unlikely. Instead, let's just catch 0.0.0.0 and
turn it into "" when we parse ports.
Fixes#7014
Signed-off-by: Matthew Heon <mheon@redhat.com>
The initial version of the new port code mistakenly restricted
this, so un-restrict it. We still need to maintain the map of
container ports, unfortunately (need to verify if the port in
question is a duplicate, for example).
Fixes#7062
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
With the advent of Podman 2.0.0 we crossed the magical barrier of go
modules. While we were able to continue importing all packages inside
of the project, the project could not be vendored anymore from the
outside.
Move the go module to new major version and change all imports to
`github.com/containers/libpod/v2`. The renaming of the imports
was done via `gomove` [1].
[1] https://github.com/KSubedi/gomove
Signed-off-by: Valentin Rothberg <rothberg@redhat.com>
I didn't believe that this was actually legal, but it looks like
it is. And, unlike our previous understanding (host port being
empty means just use container port), empty host port actually
carries the same meaning as `--expose` + `--publish-all` (that
is, assign a random host port to the given container port). This
requires a significant rework of our port handling code to handle
this new case. I don't foresee this being commonly used, so I
optimized having a fixed port number as fast path, which this
random assignment code running after the main port handling code
only if necessary.
Fixes#6806
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
Two areas needed tweaking to accomplish this: port parsing and
binding ports on the host.
Parsing is an obvious problem - we have to accomodate an IPv6
address enclosed by [] as well as a normal IPv4 address. It was
slightly complicated by the fact that we previously just counted
the number of colons in the whole port definition (a thousand
curses on whoever in the IPv6 standard body decided to reuse
colons for address separators), but did not end up being that
bad.
Libpod also (optionally) binds ports on the host to prevent their
reuse by host processes. This code was IPv4 only for TCP, and
bound to both for UDP (which I'm fairly certain is not correct,
and has been adjusted). This just needed protocols adjusted to
read "tcp4"/"tcp6" and "udp4"/"udp6" based on what we wanted to
bind to.
Fixes#5715
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
As part of this, make a major change to the type we use to
represent port mappings in SpecGen (from using existing OCICNI
structs to using our own custom one). This struct has the
advantage of supporting ranges, massively reducing traffic over
the wire for Podman commands using them (for example, the
`podman run -p 5000-6000` command will now send only one struct
instead of 1000). This struct also allows us to easily validate
which ports are in use, and which are not, which is necessary for
--expose.
Once we have parsed the ports from the new struct, we can produce
an accurate map including all currently requested ports, and use
that to determine what ports need to be exposed (some requested
exposed ports may already be included in a mapping from --publish
and will be ignored) and what open ports on the host we can map
them to.
Signed-off-by: Matthew Heon <matthew.heon@pm.me>
Paul Holzinger
Matthew Heon
Ed Santiago
Matthew Heon
Valentin Rothberg
Paul Holzinger
bitstrings
Giuseppe Scrivano
Daniel J Walsh
Luap99
baude
OpenShift Merge Robot
Adis Hamzić
5eraph
Akihiro Suda
Qi Wang