Kolla-Ansible Ceph deployment mechanism has been deprecated in Train [1].
This change removes the Ansible code and associated CI jobs.
[1]: https://review.opendev.org/669214
Change-Id: Ie2167f02ad2f525d3b0f553e2c047516acf55bc2
* Adding zuul centos-source/ubuntu-source ceph-ansible jobs
* Jobs will deploy all Ceph integrated OpenStack components, i.e.
cinder, glance, nova
* Will utilize core openstack testing script
Depends-On: https://review.opendev.org/685032
Depends-On: https://review.opendev.org/698301
Implements: blueprint ceph-ansible
Change-Id: I233082b46785f74014177f579aeac887a25b2ae2
Introduce kolla_address filter.
Introduce put_address_in_context filter.
Add AF config to vars.
Address contexts:
- raw (default): <ADDR>
- memcache: inet6:[<ADDR>]
- url: [<ADDR>]
Other changes:
globals.yml - mention just IP in comment
prechecks/port_checks (api_intf) - kolla_address handles validation
3x interface conditional (swift configs: replication/storage)
2x interface variable definition with hostname
(haproxy listens; api intf)
1x interface variable definition with hostname with bifrost exclusion
(baremetal pre-install /etc/hosts; api intf)
neutron's ml2 'overlay_ip_version' set to 6 for IPv6 on tunnel network
basic multinode source CI job for IPv6
prechecks for rabbitmq and qdrouterd use proper NSS database now
MariaDB Galera Cluster WSREP SST mariabackup workaround
(socat and IPv6)
Ceph naming workaround in CI
TODO: probably needs documenting
RabbitMQ IPv6-only proto_dist
Ceph ms switch to IPv6 mode
Remove neutron-server ml2_type_vxlan/vxlan_group setting
as it is not used (let's avoid any confusion)
and could break setups without proper multicast routing
if it started working (also IPv4-only)
haproxy upgrade checks for slaves based on ipv6 addresses
TODO:
ovs-dpdk grabs ipv4 network address (w/ prefix len / submask)
not supported, invalid by default because neutron_external has no address
No idea whether ovs-dpdk works at all atm.
ml2 for xenapi
Xen is not supported too well.
This would require working with XenAPI facts.
rp_filter setting
This would require meddling with ip6tables (there is no sysctl param).
By default nothing is dropped.
Unlikely we really need it.
ironic dnsmasq is configured IPv4-only
dnsmasq needs DHCPv6 options and testing in vivo.
KNOWN ISSUES (beyond us):
One cannot use IPv6 address to reference the image for docker like we
currently do, see: https://github.com/moby/moby/issues/39033
(docker_registry; docker API 400 - invalid reference format)
workaround: use hostname/FQDN
RabbitMQ may fail to bind to IPv6 if hostname resolves also to IPv4.
This is due to old RabbitMQ versions available in images.
IPv4 is preferred by default and may fail in the IPv6-only scenario.
This should be no problem in real life as IPv6-only is indeed IPv6-only.
Also, when new RabbitMQ (3.7.16/3.8+) makes it into images, this will
no longer be relevant as we supply all the necessary config.
See: https://github.com/rabbitmq/rabbitmq-server/pull/1982
For reliable runs, at least Ansible 2.8 is required (2.8.5 confirmed
to work well). Older Ansible versions are known to miss IPv6 addresses
in interface facts. This may affect redeploys, reconfigures and
upgrades which run after VIP address is assigned.
See: https://github.com/ansible/ansible/issues/63227
Bifrost Train does not support IPv6 deployments.
See: https://storyboard.openstack.org/#!/story/2006689
Change-Id: Ia34e6916ea4f99e9522cd2ddde03a0a4776f7e2c
Implements: blueprint ipv6-control-plane
Signed-off-by: Radosław Piliszek <radoslaw.piliszek@gmail.com>
This patch adds configs relevant to name resolution.
Change-Id: I7ebc2409e9ec0bd875abf0bf4e452bc89efe940d
Signed-off-by: Radosław Piliszek <radoslaw.piliszek@gmail.com>
VXLAN is necessary to run HA in CI (due to floating VIP
address handled by keepalived).
It also turned out to be required to have private
IPv6 address assignments.
This patch is based on linux bridge rather than OVS
to avoid problems with OVS deployed in containers.
This patch enables haproxy in multinode jobs.
Includes saving of linux networking details.
Makes DASHBOARD_URL agree with OS_AUTH_URL - properly uses the
pre-upgrade value for testing.
Co-authored-by: Radosław Piliszek <radoslaw.piliszek@gmail.com>
Depends-on: https://review.opendev.org/683068
Depends-on: https://review.opendev.org/682957
Change-Id: I66888712da80c3d6f84ee4949762961664d3adea
1) ceph-nfs (ganesha-ceph) - use NFSv4 only
This is recommended upstream.
v3 and UDP require portmapper (aka rpcbind) which we
do not want, except where Ubuntu ganesha version (2.6)
forces it by requiring enabled UDP, see [1].
The issue has been fixed in 2.8, included in CentOS.
Additionally disable v3 helper protocols and kerberos
to avoid meaningless warnings.
2) ceph-nfs (ganesha-ceph) - do not export host dbus
It is not in use. This avoids the temptation to try
handling it on host.
3) Properly handle ceph services deploy and upgrade
Upgrade runs deploy.
The order has been corrected - nfs goes after mds.
Additionally upgrade takes care of rgw for keystone
(for swift emulation).
4) Enhance ceph keyring module with error detection
Now it does not blindly try to create a keyring after
any failure. This used to hide real issue.
5) Retry ceph admin keyring update until cluster works
Reordering deployment caused issue with ceph cluster not being
fully operational before taking actions on it.
6) CI: Remove osd df from collected logs as it may hang CI
Hangs are caused by healthy MON and no healthy MGR.
A descriptive note is left in its place.
7) CI: Add 5s timeout to ceph informational commands
This decreases the timeout from the default 300s.
[1] https://review.opendev.org/669315
Change-Id: I1cf0ad10b80552f503898e723f0c4bd00a38f143
Signed-off-by: Radosław Piliszek <radoslaw.piliszek@gmail.com>
Adds four new CI jobs for testing centos/ubuntu binary/source deploys
with ironic enabled. These are run only when there are changes to the
ironic role.
Performs some simple testing by creating a node using the fake-hardware
hardware type and creating a server.
Change-Id: Ie669e57ce2af53257b4ca05f45193cb73f48827a
Depends-On: https://review.opendev.org/664011
Deploys a bifrost container using kolla-ansible bifrost-deploy.
IPA and disk images are downloaded rather than built to improve
reliability.
Currently only minimal testing of the deployment is performed, creating
and deleting an ironic node. Ideally we would perform a bare metal node
deployment.
The job is based on CentOS, as Ubuntu bifrost deployment is currently
failing with a python-MySQLdb error.
Change-Id: Ic45094594c21116b5b0d6a606f568fc7954175e3
- move check container failure from post.yml to run.yml
- add binary related jobs
- use static kolla-ansible src dir, which is helpful for kolla project
to use.
- generate correct /etc/hosts by using private ip address and hostname
- fix the wrong api interface in global.yml file
Change-Id: Idfdee6dfe18f0fa2d4f984df59b57553122ce298
Fluentd send logs to stdout,
this changes creates a file with fluentd logs
output to easy discover missing patterns during fluentd
changes.
Change-Id: I131f95089eac60ccb4c48cf5071c3b44c5ea42ca
Log retrieval was out of sync since repo split
and from multinode gates.
Many useful information retrieved before like
ps, df, docker info, etc is not in kolla-ansible
gates.
Also, his change fix logs visualization to have
colored view, allowing to easily identify errors.
Change-Id: I948233e26ceb6efc58b962bcb4b710b3f006232b
This patches changes deploy_gate quite a bit so in reality all
deployments will now assume multinode (even if it's single node). After
that we will refactor it even further to enable easy addition of new
scenerios.
Change-Id: I1faada46e6a7aa026128b2f01d77eabb04759439