Age | Commit message (Collapse) | Author | Files | Lines |
|
RDO builds packages which are aarch64 compatible but some configuration
is needed to succesfully deploy.
This change:
- Prepares the aarch64 docker.io repo as the source for Kolla Containers
- Configures VM sizing for aarch64 undercloud.
- Configures VM sizing for aarch64 virtual deploy targets.
Vms need to be larger on aarch64 compared to x86 to avoid
starvation of resources. (MYSQL)
- Configures vda2 as the location of the Linux Kernel in aarch64 in
an UEFI system
- Configures the vNICs to be on the pci-bus instead of the virtio-mmio
bus.This will enalbe the Nics to come up in the same order as the
x86 ones, so the extra configuration in ansible is not needed
- Configures apex to use a stable version of the ceph:daemon container
- Configure apex for containerized undercloud in Rocky
- Add extra ansible.cfg file for aarch64 which increases waiting
times in ansible for aarch64
- Provide helper scripts for DIB to create aarch64 UEFI images
Known limitations:
- Selinux is interfering with DHCP requests in ironic and ssh
so it must be disabled before the deploy command is ran.
- The aarch64 containers are frozen for in this commit:
https://trunk.rdoproject.org/centos7-rocky/f3/18/f3180de6439333a2813119ad4b00ef897fcd596f_70883030
- The 600s timeout defined in :
https://bugs.launchpad.net/tripleo/+bug/1789680 is not enough for
aarch64. A value of 1200s is recommended
JIRA: APEX-619
Change-Id: Ia3f067821e12bba44939bbf8c0e4676f2da70239
Signed-off-by: Charalampos Kominos <Charalampos.Kominos@enea.com>
Signed-off-by: ting wu <ting.wu@enea.com>
|
|
OOO team is removing the undercloud disk image as it is no longer needed
for containerized undercloud deployments. Instead, we can just use the
overcloud image as the undercloud image.
Additionally, OOO team has recommended we use current-tripleo instead of
current-tripleo-rdo. current-tripleo-rdo was previously thought to be
more stable with more promotion checks, but now it seems that it is
older and current-tripleo now has the same stability/checks.
This patch also bumps the undercloud RAM from 8GB to 10GB. With the new
containerized undercloud there is more RAM consumption during
deployment.
Change-Id: I9e6bb2260dbe9f8796ee54d20527c0aad96476ec
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
It looks like the docker_container ansible module will recreate the
container if it fails to restart it. This is undesired behavior so
moving to use shell to restart the containers.
Also, fixes mistral executor container not properly mounting the
ceph-ansible playbook. Additionally fixes an issue with ceph-ansible by
downgrading the package. Related rhbz:
https://bugzilla.redhat.com/show_bug.cgi?id=1644713
Change-Id: I3199b4af11a4170d19419f70cb53f7d74def273c
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Master code only supports containerized undercloud now, so this
migration is needed.
- Containerized services in undercloud
We can still apply patches to THT and other non-docker services, but
we will need to add support for patching openstack services on
undercloud.
Change-Id: I1ca4c6108f144efef7b5889503af265ef0fff8b2
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Upstream has moved to install undercloud as containers which breaks our
post undercloud configuration for some services. Disable it for now and
then move to it gracefully in a future patch.
Change-Id: I0cd1a8ddac4ba92734750265d2a16d3ef008f236
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
This patch adds capability to deploy kubernetes cluster instead of openstack.
Kubernetes will be deployed using kubespray and is run after TripleO bookstraps
overcloud nodes.
JIRA: APEX-574
Change-Id: If9c171620c933a052b719e7112a50e22bbab667f
Signed-off-by: Feng Pan <fpan@redhat.com>
Signed-off-by: Zenghui Shi <zshi@redhat.com>
|
|
In latest THT upstream, environment file neutron-opendaylight-sriov.yaml was
moved to services folder. Updating references in Apex to avoid deploy failure.
Change-Id: I7065e0d8e13c9add9ead282db2244a27c177e5a4
Signed-off-by: Feng Pan <fpan@redhat.com>
|
|
|
|
Baremetal deployments were failing because the ceph PG size was
exceeding the max allowed. Virtual was still working because we lower
the number of pools and pg/osd. This patch changes the values to a
number which should work for both virtual and baremetal. Also includes a
fix which adds the controllers back as OSDs and a few other cleanup
issues.
JIRA: APEX-614
JIRA: APEX-569
Change-Id: I2ad65727ecdcaa0454eb53d25e32b7f1a53cd3a4
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
/var/lib/mistral path contains logs for when ansible is invoked by
TripleO for Ceph configuration as well as config download. This patch
now archives and fetches that directory. Logs in previous releases
like Queens store the Ceph logs in /var/log/mistral.
Change-Id: I50c43e55efaa5dbcf8b7fb00b0e11cd3288fdd05
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
There was a compatibility issue with the centos 7.4/7.5 between the host
pacemaker version and container. Now that containers have moved to 7.5
we should not need this workaround anymore.
Change-Id: I9632c65e87687d4f36130719c6df9af2e913eed8
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
We now move master to deploy from upstream. That means we do not need
to build undercloud/overcloud images anymore.
Changes-Include:
- Remove bash build scripts as we do not need to build anything
other than the python package anymore
- Remove building images or iso from build.py
- Remove building of images and iso from Makefile
- Rename/refactor deploy settings files for nosdn and odl. The new
convention is that the typical scenario names we use will deploy
master. We also support n-1 OS, so in that case we use the branch
name for the "feature" in the scenario name: os-odl-queens-noha.
- Tacker/Congress are disabled in settings files until we fix that with
upstream. Containers are now enabled by default.
- Disable TLS for undercloud (was changed upstream to default enabled)
- Fix environments docker directory for master THT (was changed upstream)
- Includes fix for LP#1768901
- Includes workaround for LP#1770692
- Moves to docker.io for container images as it is more stable and
should contain the same images
- Removes the term 'common' from apex packaging for referencing the
Python Apex package
Change-Id: If6b433860b3ff882686c78d0f24a2f0c52b9b57a
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
This scenario should enable SRIOV interfaces to be used
by Neutron. Only will be supported in baremetal deployments
with SRIOV capable NICs. The name of the interface must
be known in advance and the physnet of the SRIOV network
is set as nfv_sriov.
Change-Id: Ie4295413e0be2197bd9ada4f887f6b47cd486765
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
|
|
- exposes new option to end users to skip introspection
- moves the logic to decide to introspect or not into python
JIRA: APEX-536
Change-Id: Ieaff11362ff8f906daa98d301d3d473ad549d08f
Signed-off-by: Dan Radez <dradez@redhat.com>
|
|
The lineinfile was not actually inserting the CephOSD line because it
was already present later in the file (regardless of 'insertbefore').
Therefore we can fix it by simply removing the CephOSD line before we
try to set it, since we do not need it in the storage role anyway.
Change-Id: I8c8d9b6baccfc77ea582fab6ad438b02293f96fb
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
There was an issue with patching the overcloud where the patch binary is
missing, making it impossible to apply patches. This change install
patch now on the image.
Also, although deployments were successful, storage was not working.
This is because by default upstream does not apply Ceph OSDs to compute
nodes for hyperconverged Ceph, but we use this as our standard
deployment in Apex. This patch inserts CephOSD into the default Compute
role. Note: we normally override role's services in regular Apex
deployments so we do not hit this issue there.
Change-Id: I5bddda4784dc00148395863ae0990343a4159602
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
To deploy with upstream openstack branch, use new deploy setting 'os_version'.
A default scenario file for nosdn with pike has been included in this patch.
If 'os_version' is a version other than the default version for this OPNFV
release, then upstream is used.
In order to use upstream with the current OS version use '--upstream'
argument to the deploy command, to force an upstream deployment.
Also include '-e upstream-environment.yaml' to use default upstream
deployment settings.
Supports nosdn and odl-nofeature deployments.
Change-Id: Ic07e308827b449637b4e86cdd086434e4de2fb69
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
Change-Id: I630e90065f9cc76ef42ca9b23aea920c2493b53f
Signed-off-by: Dan Radez <dradez@redhat.com>
|
|
Change-Id: Ic09a9b6d57d2ab05562f64a75f9732f76ac61bb7
Signed-off-by: Tim Rozet <trozet@redhat.com>
|
|
- move patches from x86 only to common virt-customize
- updates to undercloud to configure pxe and external network
- initial updates to overcloud deploy functions
JIRA: APEX-381
Change-Id: I496e196535e63bfd31654d0f501fe3f00fa5ca3f
Signed-off-by: Dan Radez <dradez@redhat.com>
|
|
Removes all bash libraries and converts almost all of the code to a
mixture of Python and Ansible. utils.sh and clean.sh still exist.
clean.sh will be migrated fully to clean.py in another patch.
The Apex Python package is now built into the opnfv-apex-common RPM. To
install locally do 'pip3 install .'. To deploy:
opnfv-deploy -d <file> -n <file> --image-dir /root/apex/.build -v --debug
Non-python files (THT yaml, settings files, ansible playbooks) are all
installed into /usr/share/opnfv-apex/. The RPM will copy settings files
into /etc/opnfv-apex/.
JIRA: APEX-317
Change-Id: I3232f0329bcd13bce5a28da6a8c9c84d0b048024
Signed-off-by: Tim Rozet <trozet@redhat.com>
|