We're moving this because we discovered that Octavia will not always
raise Conflict if you create an LB with the same IP.
If no VIP port got created for an LB, then Octavia will not have any
issue with creating another LB with the same IP.
Closes-Bug: 1947809
Change-Id: I26b911b30403c7ea5a35706b2fc80af499e91330
When the Pod has status Completed it has finalized
its job, consequently it's expected that Kuryr will
reclycle the Neutron Ports associated with it to be
used by other Pods. However, Kuryr is considering
Completed Pods as not scheduled and the event is skipped.
This commit fixes the issue by ensuring that Completed
Pods Ports recycle is done before checking for Scheduled
Pods.
Closes-Bug: #1945680
Change-Id: I23e7901fb016d59fa76762d1f24fee47974e72df
let containers aware of the host's netns changes
Signed-off-by: yue.lg <yuelg@chinaunicom.cn>
Change-Id: I312817db503002f918276d0ff41baeb816f39404
Closes-Bug: 1945293
For now, by default, pod network was taken from OpenShift Machine spec,
by looking for "networks" field and get the first item out of it to
figure out right subnet. Currently there is another field, where
networks can be defined - under "ports" field, where subnet for fixed
IPs can be found. Additionally, to avoid confusion, there is another
field, which defines primary subnet for the pods - "primarySubnet".
Implementation has been adapted to reflect that change:
1. primarySubnet will be examined, to get the subnet id
2. if primarySubnet does not exists, networks field will be examined
3. if there is no networks field (or it doesn't contain subnets), ports
field will be examined for subnet id
Change-Id: I35b6f2e6288cbe3fd80b74a5da03639d7bd0a765
In Kuryr-kubernetes we mainly rely on two services: Neutron for ports
creating and Octavia for load balancers. Sometimes we observe that either
ports or load balancer hangs in some states indefinitely. In such case,
after we timeout, kuryr controller will crash, leaving ambiguous log
message.
In this patch we introduce two new Prometheus metrics - one for Port and
the other for load balancer, which will be updated during such
situation.
Change-Id: Ifa4792019be97c21d1d531d9409f95ffd6d54499
Includes a new Gauge metric that records the
number of members of load balancers considered
critical. The metric is labeled with the Name
of load balancer and pool name, and the amount
of members. Also includes an Enum with the
current state of the lb.
Change-Id: Id89bb48d86588f4d2a28ab91963e0b84843cbd6f
We don't need to log every successful /metrics call on INFO level. This
commit switches the log level there to WARNING.
Change-Id: I3bc0e1c9e312381715f60a00ad8f52dbd601bb2f
Seems like a thread moving the metrics data into the service exposing
them in the Prometheus format was not constrained with any sleep. This
was causing increased CPU usage of kuryr-cni pods without a real
reason to do it.
This commit solves that by rewriting the thread to use the common
multiprocessing queue pattern.
Change-Id: I0eacc37022fbf214c361dbc52b42281ffa5301fd
on the line #114 we set the liveness probe to False, which triggers
restart of the Kuryr Controller.
However without DEBUG, the LOG message is hidden behind DEBUG.
This makes it harder to troubleshoot if the Handler failed with this
exception
Change-Id: I71deb8434273ce39453d69a05d6bd8b105fe2c2b
- update grpcio in requirements.txt
- update cffi, greenlet and grpcio in l-c to avoid problems
when trying to build the wheels.
- adds .stestr to gitignore
Change-Id: Id2d1c3ea60c33803a35082b9064ccba982b014a7
We had restricted version for <0.8 and >0.7.2 (which is 8 years old
btw), which probably breaks the interface, although current version,
which is 2.2.1 works just fine with kuryr-status command. With this
patch, we allow to install latest version, so that it will not break
requirements-check test.
Change-Id: I5542db6829e01b4767fe1df11b421446bc26b6a3
It would be beneficial for the project to have it's own sanity check
command. Where we can verify the configuration used for kuryr is
supported.
Change-Id: I1a11694e938e00be653697e8fd96e071faacc96a
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 prevents Kuryr Controller from CrashLoopBackoff in unexpected
case where the port is still part of the trunk during _cleanup_leftover_ports
Change-Id: Ie3f75118a58094d6ca4287e1ffc0035d380d0584
Kubernetes provides a convenient way for examining resource life cycle.
Ephemeral Event objects are created for some of the steps during
resource object change of the state. Those event can be accessed using
`kubectl get events` or for particular object `kubectl describe <type>
<resource name>`, so that operator can examine the state and the reason
for potential failure of given resource object.
In this patch there is a method added for creating Event objects, so
that anything wrong regarding Kubernetes object life cycle can be
"annotated" with such object.
Change-Id: I291613adf9282ec5399fbe5de75e82472c2bc9c0
This patch is originally an automatically generated patch to ensure
unit testing is in place for all the of the tested runtimes for the
next given cycle (this time yoga). However there was an error during
the patch generation as a gitignore rule excludes all hidden directory.
This patch is now removes the general hidden directory ignore rule,
too, to avoid such errors in the future.
Change-Id: I3341f08a7f331fb8b97b2d75106eefcec9b2d5d7
Add file to the reno documentation build to show release notes for
stable/xena.
Use pbr instruction to increment the minor version number
automatically so that master versions are higher than the versions on
stable/xena.
Sem-Ver: feature
Change-Id: Ief61397c9323e3504de07051a2279afc116f023b
Octavia tempest plugin is required to test for loadbalancer
reconciliation. As a result, there is need to configure the
gate to enable reconciliation by default.
Implements: blueprint reconcile-openstack-resources-with-k8s
Change-Id: Ifab3a4f45da29141d7889bc0f110ace6c5b96eab
This change adds help strings for cache parameters, so that description
of these parameters are included in .conf file generated by
oslo-config-generator.
Change-Id: I9d7082fc42b5dbd67f7214810719158d93f5f88d
The patch handles the KuryrLoadBalancers reconcilation
using periodic task as event scheduler instead of eventlet
Implements: blueprint reconcile-openstack-resources-with-k8s
Change-Id: I85d2d1cf00b2543f41a875b32ae64fbbe885652c
As part of the change below, we are adding a utility code to get klb crd
path.
The klb crd shares the same namespace and the name as the service,
endpoints.
This code can be hence re-used for services too.
Depends-On: https://review.opendev.org/c/openstack/kuryr-kubernetes/+/804205
Change-Id: I90c8851e01de16b82617d60c987bad50a1522dd5
Zuul
Robin Cernin
jayonlau
Maysa Macedo
yue.lg
Roman Dobosz
gryf
Michał Dulko
Maysa De Macedo Souza
Luis Tomas Bolivar
Iury Gregory Melo Ferreira
Elod Illes
OpenStack Release Bot
Sunday Mgbogu
Takashi Kajinami