| Age | Commit message (Collapse) | Author |
|---|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
With recent changes to oe-core make clean is run during the
configuration phase of python packages.
For some packages, this causes breakage as 'make clean' is not
supported.
To keep the build going, we mark them as brokenclean to avoid
the issue.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
oe-core now warns on detected missing runtime and build time warnings.
So we update our recipes to have these missingn deps (largely bash).
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Updating the bbappends to match the oe-core versions ~commit 530565c6cb77319f4cc49edeea6fec6be997415d
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Include monitoring packagegroups into various openstack
packagegroups.
Monitoring feature can be turned on by including the following
in local.conf
OPENSTACK_EXTRA_FEATURES += "monitoring"
There are various different system monitoring recipes
(e.g. Nagios or Monasca) provide packagegroup-monitoring-xxxxx.
To choose what system monitoring to use set the following in
local.conf
PREFERRED_PROVIDER_virtual/monitoring = "packagegroup-nagios-monitoring"
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
Instead of having a central file or group of files to
describe what data resources should be monitored. The
content of these files will depend on what core system
monitoring is used ((e.g. Nagios or Monasca).
It's desirable to have each recipe describes what
it wants be monitored in generic way such that various system
monitors can understand and convert these into their format.
If a recipe wishes to register itself to system monitor, it
inherits monitor bbclass and use MONITOR_SERVICE_PACKAGES and
MONITOR_SERVICE_<package name> to indicate what processes
should should be monitored. Also MONITOR_CHECKS_<package name>
variale can be used to pass list of scripts which will be run
on target and if any of these scripts fail then will report.
Eventually monitor.bbclass will be expanded to allow recipe
to describe more complicated information passed down to
system monitor (e.g. Nagios or Monasca)
Signed-off-by: Vu Tran <vu.tran@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Although the use of PRINC is deprecated in later versions of Yocto it
may still be used and if you are using this layer with older Yocto it
is recommended for use in bbappends. It is therefore expected to
work. PRINC expects a non-numeric prefix followed by a numeric value,
as can be seen in base.bbclass
pr_prefix = re.search("\D+",pr)
prval = re.search("\d+",pr)
if pr_prefix is None or prval is None:
bb.error("Unable to analyse format of PR variable: %s" % pr)
Failing to stick to this convention yields a parsing error when you
attempt to use PRINC:
ERROR: Unable to analyse format of PR variable
Adding the non-numeric prefix allows PRINC use in bbappends to
function correctly.
Signed-off-by: Mark Asselstine <mark.asselstine@windriver.com>
|
|
Openstackchef class provides the framework to reconfigure openstack
at run-time. These services are inheriting openstackchef so that
script file(s), specified in the respective recipe file,
using the variables below are restarted by the decentralized
openstackchef class framework when reconfiguring openstack:
INITSCRIPT_PACKAGES: The list of init-scripts to start/stop when
reconfiguring openstack.
INITSCIPT_NAME_x: The name of the init-script mentioned above.
INITSCRIPT_PARAMS_X: This parameter defines the run-level and
priority for the init-script. However, the only parameter in this
variable that is of interest to openstackchef class is the priority of
the init-script.
Changes to none of the above variables is shown by this patch set,
however, it's to point out that openstackchef makes the assumption
that these variables are set in the service's recipe file.
Failure to set these variables, could lead to the service not
working properly after the stack is re-configured.
Signed-off-by: Mustapha Lansana <Mustapha.Lansana@windriver.com>
|
|
Openstackchef enables us to recreate configuration files
for services in an openstack installation. It does this by
creating template file(s) out of configuration file(s) exposed
to the class by services.
The following services are inheriting the openstack class
and then exposing a set of configuration files to the class.
These services expose their configuration files to openstackchef
by assigning them to the variable CHEF_SERVICES_CONF_FILES. The files
are assumend to have been installed in the image directory under the
service's WORKDIR.
At build-time, openstackchef makes chef-solo templates out of
the registered files. And at run-time, the deploychef package
makes a call to chef-solo, which in-turn use the template files
to recreate the registered configuration files.
For legacy reasons, the string OPENSTACKCHEF_ENABLED is defined in
openstackchef class, but it can be overwritten in a .bb, .class,
.bbappend or local.conf file to an empty string when openstackchef
support is not desired. This enables all of these services to be built
without openstackchef support. In addition, it prevents the recipes
from substituting the placeholders in their configuration files
when inheriting openstackchef.
Signed-off-by: Mustapha Lansana <Mustapha.Lansana@windriver.com>
|
|
"task-ceph" package group is renamed to "packagegroup-ceph"
so update any file in meta-cloud-services layer that uses
"task-ceph"
Signed-off-by: Vu Tran <vu.tran@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
This commit changes all required configurations to use keystone running
on apache. The following packages configurations were modified for
keystone running on apache: python-neutron, python-nova, tempest,
python-swift, python-rally, python-heat, python-glance,
python-cinder, python-ceilmoeter, python-horizon.
Signed-off-by: Liam R. Howlett <Liam.Howlett@WindRiver.com>
|
|
Update troveclient to newer git commit to support postgresql and split
the package into a meta-package, the python package, bin package, and
test package. Add the meta-package to the compute image & trove RDEPENDS.
Signed-off-by: Liam R. Howlett <Liam.Howlett@WindRiver.com>
|
|
Create the OpenStack Database as a service recipe and add it to the
controller image. The recipe includes the init script and initial setup
to use postgres.
Signed-off-by: Liam R. Howlett <Liam.Howlett@WindRiver.com>
|
|
Removing PRINC from recipes to silence the bitbake warning about
it being a depreciated variable.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
when launching guests, libvirt may create veth links, without having
the veth module on the target this fails, and no guests can be launched.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Amy Fong <amy.fong@windriver.com>
|
|
Instead of creating tenant/user/role and service/endpoint for all
openstack services in keystone postinstall, now each of the services
creates its own keystone identities by queueing them up in its postinstall
to a file /etc/keystone/service-user-setup. service-user-setup
script, when run as the last postinstall, calls identity.sh with keystone
identity parameters to create necessary identities for the services.
Signed-off-by: Andy Ning <andy.ning@windriver.com>
|
|
- modify pam configuration files to use ldap
- modify sshd to enable pam
- modify nsswitch.conf to use ldap
Signed-off-by: Amy Fong <amy.fong@windriver.com>
|
|
Enabling OpenLDAP support in keystone and the controller as a distro
feature
Add packages openldap python-ldap nss-pam-ldapd, libsasl2-modules,
python-keystone-hybrid-backend, pam-plugin-mkhomedir and cyrus-sasl to
the controller task
Signed-off-by: Amy Fong <amy.fong@windriver.com>
|
|
Include Rally benchmarking feature into all-in-one
image when OPENSTACK_EXTRA_FEATURES variable
includes "benchmarking" feature.
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
uwsgi is using pkg-config to detect whether or not yajl is present
in the build. We do want to use yajl, so we add an explicit dependency
and pkgconfig inherit.
Without this, build errors may occur, since the build paths that don't
use yajl are not supported.
And finally, to prevent the searching of host include paths, we set
UWSGI_REMOVE_INCLUDES and purge local directories. This ensures that
only the sysroot and package config settings are used.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Since rabbitmq uses HOME as the location for the erlang cookie,
rabbitmq will fail to start if that directory is not on the
system.
To ensure that this is configurable, we use the bitbake variable
ROOT_HOME and replace a placeholder in the initscript.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Create new benchmarking packagegroup bb. Benchmarking
packagegroup can be included into final image to include
all required packages for benchmarking OpenStack at scale.
Currently OpenStack benchmarking is provided by OpenStack
component Rally.
Also introducing new Yocto variable OPENSTACK_EXTRA_FEATURES
which can be used to inform what feature should be included in
meta-cloud-services OpenStack final image.
To include benchmarking packagegroup in local.conf have
OPENSTACK_EXTRA_FEATURES += " benchmarking"
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
OpenStack Rally requires "tools" from tempest.
Also modify tempest ".testr.conf" to allow testr to
run tempest from any location.
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
On system bootup, the softdog kernel module gets modprobed from the package
sanlock-wdmd-init, but because it didn't exist, we get:
Loading the softdog kernel module: FATAL: Module softdog not found.
Failure
Add the package kernel-module-softdog to add the kernel module softdog
to the rootfs
Signed-off-by: Amy Fong <amy.fong@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
The metadata service is working as the following:
- metadata is being served by nova-api on controller at port 8775.
- VM instance requests metadata by 169.254.169.254 (eg, curl http://169.254.169.254/latest/meta-data)
- metadata request comes to neutron-ns-metadata-proxy on controller in dhcp network name space.
- neutron-ns-metadata-proxy forwards the request to neutron-metadata-agent through a unix domain socket (/var/lib/neutron/metadata_proxy).
- neutron-metadata-agent sends the request to nova-api on port 8775 to be serviced.
To support metadata service, neutron-ns-metadata-proxy is baked into the controller image. Also neutron-metadata-agent startup script
(/etc/init.d/neutron-metadata-agent) and config file (/etc/neutron/metadata_agent.ini) are added to start up metadata agent at system
initialization. dhcp_agent.ini and nova.conf are updated as well.
A README.metadata is added in the Documentation/ directory.
Signed-off-by: Andy Ning <andy.ning@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
cirros 0.3.0 has a bug which ignores classless static routes passed by
dhcp option 121. This static route is needed for the instance to access
metadata service from 169.254.169.254. 0.3.2 has this bug fixed.
Details about this bug can be found at: https://bugs.launchpad.net/cirros/+bug/1190372
Updated documents that reference specific version of cirros as well.
Note: 0.3.2 download URI (SRC_URI) has been changed from "https://launchpad.net/cirros/trunk" to "http://download.cirros-cloud.net", since cirros binary downloads are all in the new URI.
Signed-off-by: Andy Ning <andy.ning@windriver.com>
|
|
* Build failed with this error:
| x86-64-wrswrap-linux-gnu-g++ -m64 --sysroot=/media/vtran/1-build/build_temp/build-ovp-1/bitbake_build/tmp/sysroots/intel-xeon-core -O2 -pipe -g -fvisibility-inlines-hidden -c -o gptpart.o gptpart.cc
| In file included from guid.cc:22:0:
| guid.h:29:23: fatal error: uuid/uuid.h: No such file or directory
due to missing "util-linux" in DEPENDS.
* Also got this warning:
WARNING: QA Issue: gptfdisk: /sbin/cgdisk, installed in the base_prefix, requires a shared library under exec_prefix (/usr): libstdc++.so.6 => /usr/lib64/libstdc++.so.6 (0x00000000dead4000)
Move binaries to /usr/sbin/ instead of /sbin/
Signed-off-by: Vu Tran <vu.tran@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
The setup.py script that is included with the cloud-init code
is setup to install files into the "/usr/lib" directory. On
64-bit machines the target directory should be "/usr/lib64".
This generates a warning when using bitbake. This fix performs
a search and replace on the setup.py file to set the correct
output directory. This fix also adds a symbolic link to each
script from the /etc/cloud directory to make the scripts
available from a more obvious location.
Signed-off-by: Keith Holman <Keith.Holman@windriver.com>
|
|
When populating the Openstack rootfs with rabbitmq-server package,
you get the following warnings:
WARNING: QA Issue: rabbitmq-server: Found a reference to /usr/ in
rabbitmq-server-3.2.4-r2/packages-split/rabbitmq-server/sbin/rabbitmq-env
WARNING: QA Issue: Shell scripts in base_bindir and base_sbindir
should not reference anything in exec_prefix
This patch silence the warning by telling the build system to not
throw warnings if a script in base_[bindir | sbindir | libdir ] makes
what it considers an unsafe reference to exec_prefix.
The reference though to exec_prefix in this case is very safe, since
it is only part of variable directing other scripts to
rabbitmq-server's plugins directory.
Signed-off-by: Mustapha Lansana <Mustapha.Lansana@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
GPT fdisk (gptfdisk) is a disk partitioning tool loosely modeled on
Linux fdisk, but used for modifying GUID Partition Table (GPT) disks.
When prepare disk device for ceph with ceph-disk, it's required sgdisk
util from gptfdisk to perform disk partition. So add gptfdisk.bb.
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
uwsgi needs uuid/uuid.h to be present, otherwise we get the following
build error:
| In file included from core/utils.c:1:0:
| ./uwsgi.h:188:23: fatal error: uuid/uuid.h: No such file or directory
| #include <uuid/uuid.h>
| ^
| compilation terminated.
Adding a e2fsprogs dependency gets the header file properly installed before
the build starts.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Some OpenStack components expect the system to contain a
syslog user for logging purposes and fail if one doesn't
exist. This fix adds the syslog user and group within the
sysklogd package since it responsible for the system
logging capabilities of the system. This fix also adds a
dependency between the package requiring the syslog user
to exist.
Signed-off-by: Keith Holman <Keith.Holman@windriver.com>
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
The tempest fixes to prevent an explicit version from being added
to urls was developed for havana and doesn't directly apply to
icehouse. Removing the patches, and they will be examined in the
future to see if they are applicable.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Currently glance keystone endpoints URLs include
"v2/" at the end (e.g. http://<glance server ip>:9292:v2/).
This means glance should only be talked to using v2.
For tempest image testcases, image_client.py gets URLs
from keystones and appends additional version into
these URL strings (e.g. htt://<glance server ip>:9292:v2/v2/images)
which causes glance not to understand the command and return error:
NotFound: Object not found
Details: 404 Not Found
The resource could not be found.
In our case, we use the "v2" from URLs.
We also disable v1 image testcases as it's not possible
to test v1 with current glance endpoind URLs.
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
commit 2b5287db8116ef8e3ed5e4fc211296e6293b5dcc upstream
https://github.com/openstack/tempest.git
This commit switches the image api tests from auto detecting which
api versions are available to having them explicitly set in the config
file. This is to make it explicit which tests are expected to be run
instead of assuming that everything is expected to work.
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
As now Openstack Heat component is enabled,
so allow tempest to test Heat as well.
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
As cinder-volume supports multiple backend drivers:
nfs, glusterfs, ceph, and lvm-iscsi, it's good
to allow tempest to test cinder-volume multi-backend.
Signed-off-by: Vu Tran <vu.tran@windriver.com>
|
|
shipped in any package
A number of license and installation files were installed to ${D}, move these to
${D}/usr/share/doc
Signed-off-by: Amy Fong <amy.fong@windriver.com>
|
|
meta-virtualization versions
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Track the update of libvirt to 1.2.4 in meta-virtualization. We also move
the 1.2.2 bbappend to dangling for layers and users that are not tracking
the latest meta-virt.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
tempest no longer branches per-release and instead ensures that the
single branch works for all releases.
As such, we change to master and update the SRCREV to the latest.
There is one new binary part of the package, so we add it to the
default tempest package.
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
|
Signed-off-by: Bruce Ashfield <bruce.ashfield@windriver.com>
|
Bruce Ashfield
Vu Tran
Mark Asselstine
Mustapha Lansana
Liam R. Howlett
Amy Fong
Andy Ning
Keith Holman