DevStack stopped accepting Ubuntu 20.04 Focal, so we need to bump the
version we use to Jammy. This commit does so.
Change-Id: I7b227dfb29b49ff6a561209938a2c263f5dd2f90
For some reason linuxbridge was set as mechanism driver in OVS gates.
This commit disables that as linuxbridge became experimental in Neutron
and gates fail.
Change-Id: I2dd5c7f19af73badae49d2509efb0016335e6c9e
We face general instability of the Amphora jobs that we have. A while
ago we've made sure that these jobs are scheduled on the clouds that
support nested virtualization in order to boost performance of the
Amphora LB VMs. Now it turns out we haven't configured DevStack to
actually use kvm instead of qemu, so it's possible that we were still
running these jobs with software virtualization.
This commit attempts to reconfigure DevStack to actually use kvm and
nested virtualization.
Change-Id: I57f70c1cb140d4fa54d13b6c759a0179d8adb7b4
According to [1], the needed version for OVN v21.06.0 is
a4b04276ab5934d087669ff2d191a23931335c87
[1]4b0a02635f
Related-Bug: #1967472
Change-Id: I73a11d4a02c811366c177fa31d02599e247fea3a
There is a issue[1] with the old versions we were using,
let's attempt to bump them.
[1] https://bugs.launchpad.net/neutron/+bug/1964339
Change-Id: I4a9a9f877c2a0aea2bf3e43e179afedb653b8464
In I950e81d3f9f23493f26d9f824d3052ed375ae62b I've forced the jobs
explicitly saying "amphora" to run on clouds with nested virtualization
enabled. I however omitted the jobs that are not having "amphora" in the
name - defaults, IPv6 and OVS jobs. This commit fixes this.
Change-Id: I1e70118bb9c7d80e78806c3319c95af7efd01a08
Also, rearrange tests files.
Job hierarchy looks as follows:
- kuryr-kubernetes-base (base tempest job)
- kuryr-kubernetes-base-ovn (derived from base for neutron OVN)
- kuryr-kubernetes-tempest-amphora
- kuryr-kubernetes-octavia-base (ovn-octavia-provider)
- kuryr-kubernetes-tempest (alias for the parent,
used in projects kuryr-tempest-plugin and os-vif)
- kuryr-kubernetes-tempest-ovn-provider-ovn (alias for the parent,
used in project ovn-octavia-provider)
- kuryr-kubernetes-tempest-systemd
- kuryr-kubernetes-tempest-centos-7 (this one is probably
broken)
- kuryr-kubernetes-tempest-defaults (default settings for
kuryr-kubernetes, so OVN + amphora + default set of handlers,
default subnet and sg drivers)
- kuryr-kubernetes-tempest-ipv6
- kuryr-kubernetes-tempest-dual-stack
- kuryr-kubernetes-tempest-lower-constraints
- kuryr-kubernetes-tempest-l2
- kuryr-kubernetes-tempest-pools-namespace
- kuryr-kubernetes-tempest-crio (currently broken)
- kuryr-kubernetes-tempest-multinode
- kuryr-kubernetes-tempest-multinode-ha
- kuryr-kubernetes-base-ovs (derived from base for neutron OVS)
- kuryr-kubernetes-octavia-base-ovs (amphora)
- kuryr-kubernetes-tempest-ipv6-ovs
- kuryr-kubernetes-tempest-amphora-ovs
- kuryr-kubernetes-tempest-multinode-ovs
- kuryr-kubernetes-e2e-np (base devstack job)
For convenience, tempest jobs are in two files: one for multinode jobs,
and one for one node jobs. Network policy job is in its own file, since
it have different parent (devstack).
Change-Id: I198f138feb9bd7b6cc659ac8cd788cddd5be9439
This patch provides an implementation for joining node to the Kubernetes
cluster by using `kubeadm join` command.
Change-Id: I71d2b99e0c92a12c4e64395f6c4dafa4b69f168f
Depends-On: Ife21874c0a71ba07723094c0f880aabcf5825b77
Till now, for installing kuryr-kubernetes and one of the crucial service
- kubernetes, there has been used manual method for installing it in
specified version. Over time it became a burden to follow requirements
and constraints, therefore decision has been made to use recommended way
of installing Kubernetes - kubeadm. In this patch devstack installation
of the kuryr-kubernetes and its dependences has been heavily reworked.
Other than that, OpenShift related functions has been removed, since
they were all outdated and non-working for the long time.
Change-Id: Ife21874c0a71ba07723094c0f880aabcf5825b77
The Neutron community is planning on switching the default network
backend in DevStack to OVN soon and to avoid any gate breakages we need
to explicitly enable ML2/OVS where it makes sense.
This patch is enabling ML2/OVS for the non-OVN jobs in the gate. Prior
to this patch the job were enabling the ML2/OVS services such as q-agt
and q-dhcp but it wasn't setting other required options (e.g Q_AGENT).
Signed-off-by: Lucas Alvares Gomes <lucasagomes@gmail.com>
Change-Id: Iae420403d3d84325830bded8633b2a29430b36bb
test_network_policy_add_remove_pod is often not finishing in 20 minutes
in the gate, effectively timing out the tempest default timeout. This
commit increases the timeout to 30 minutes to avoid that.
Change-Id: I3bbac8e90553be9424fa0ecdb4c05e096da8f2d7
Apparently dstat's bug causing it to show that kuryr-daemon is using
1 TB of RAM causes some buffer overflow in Python 3.8. We need to
disable it in order to enable testing on Ubuntu Focal.
Change-Id: Ifacc8c5f3e4a7adc010de75923d69f1df615450a
This fixes 3 issues in order to unblock CI:
1. OVN job is fixed by updating it's configs to match what Neutron uses
to configure DevStack with OVN.
2. lower-constraints job will run on Ubuntu Bionic and Python 3.6,
allowing it to work with deps that don't support 3.8.
3. Swap size in our jobs is set to 8 GBs as recently the default got
shrinked to 1 GB which is way too little for our Amphora jobs.
Also some additional OVN debugging info will get gathered.
Change-Id: I70c867ac21004586c014e9eb797dbf528dd6e3f2
Per request from infra cloud providers we need to disable world-facing
DNS that is the coredns and find a way to run it privately on the gate
VMs. Which won't be that easy, as there's unbind bound to 127.0.0.1:53,
but let's see if we'll be able to come up with something.
Meanwhile this change shouldn't hurt too much as we were yet to utilize
the coredns deployment and I think it wasn't even working on the gate.
Change-Id: I0c671b533f2babef172b46c36f62fba929d724c5
Ussuri release is the one in which we drop Python 2 support, as its EOL
is pretty close now. This commit does so in kuryr-kubernetes by
removing Python 2 unit test jobs, switching all tempest jobs to Python
3, removing specific jobs for Python 3 and updating Dockerfiles to
centos:8 that includes Python 3 from the box.
Also CentOS 7 job is removed from check queue as it seems it doesn't
play well with Python 3. A CentOS 8 job will get created soon.
Change-Id: Id9983d2fd83cef89e3198b2760816cf4a851008b
Looks like Ubuntu's Amphora is broken, probably due to latest update of
ubuntu-minimal cloud image (which includes cloud-init version bump that
might be the culprit). This commit switches our gates to CentOS amp to
work the problem around until we fix the Ubuntu one.
Also the timeout for the jobs is raised to 3 hours because it seems like
building containers in the gate take ages.
Closes-Bug: 1822862
Change-Id: I3650aac01334e924ca64fcc4e79928141e010e5a
As a step to improve testing capabilities of our gates, this commit
enhances DevStack with support for deploying coredns in our K8s cluster.
The idea here is to be able to run any tests that are referring to
services by <namespace>.<service-name>, in particular upstream K8s
tests.
The tricky part here is that on gate VM's an instance of unbound DNS is
running on 127.0.0.1:53. As in DevStack-deployed Kuryr pods doesn't
support IPv6, we couldn't just take IPv6 addresses of upstream DNS from
unbound configuration and use them in coredns pods. Instead the coredns
instance is running on host networking and binds to $HOST_IP:53, which
is also used as value of kubelet's --cluster-dns option, while
forwarding any upstream DNS requests to the local unbound instance. This
isn't perfectly how it would be set up in production environment, but
should be close enough for our purposes.
This change only affects DevStack, so it's completely safe from release
point of view. coredns gets enabled only on gates running Kubernetes as
OpenShift gates run openshift-dns already.
Change-Id: Icdab52a6229b2209f58e26e4d885f551883727b5
Partial-Implements: blueprint k8s-upstream-tests
As Octavia is now the only supported LBaaS option, we can remove
"-octavia" suffix from job names. Also neutron-lbaasv2 is removed from
required projects in base job definition.
Change-Id: I35abbb73249dd907b3b5055287624929b121afb3
Implements: blueprint remove-lbaasv2
Deploying without kuryr-daemon is deprecated since Rocky and we
announced that it will be removed in the Rocky release notes. This
commit removes all the code that allows that, updates the documentation,
DevStack plugin and gates definitions.
Implements: blueprint remove-non-daemon
Change-Id: I65598d4a6ecb5c3dfde04dc5fefd7b02fc72a0cb
Option listed above will make the data directory of etcd deployed by
DevStack to be a ramfs RAM disk, which should greatly help with
Kubernetes instability by reducing fsync durations.
Change-Id: I9543c1b7aa41e10b64996b0d705593521f46b8b2
Depends-On: https://review.openstack.org/#/c/626885
Using TEMPEST_PLUGINS only installs the tempest code but does
not run the tempest devstack plugin. The tempest devstack plugin
is where we build kuryr/demo from the rootfs in the repo. If this
is not run, we are depending on dockerhub and potentially running
an older image.
Depends-On: Idf2c781591f938a02991af4b7991da19a92e03f2
Change-Id: I2cb58ab038a619f67462271e2db3ed20e9e51ba3
Signed-off-by: Antoni Segura Puimedon <celebdor@gmail.com>
This commit splits .zuul.yaml into smaller yaml files inside .zuul.d
repo for easier handling.
Partially-Implements: blueprint zuulv3-native-gates
Change-Id: I73ac4381eaaf14d515eb45c7020881247428e33a
Zuul