| Age | Commit message (Collapse) | Author |
|---|
|
Ubuntu images on IBMCloud for 16.04 have some seed data in
/var/lib/cloud/data/seed/nocloud-net. In order to have systems with
IBMCloud enabled, we modified ds-identify detection to skip that seed
if the system was on IBMCloud. That change did not consider the
fact that IBMCloud might not be in the datasource list.
There was similar logic in the ConfigDrive datasource in ds-identify
and the datasource itself.
Config drive is now updated to only check and avoid IBMCloud if IBMCloud
is enabled. The check in ds-identify for nocloud was dropped. If a
user provides a nocloud seed on IBMCloud, then that can be used.
This means that systems running Xenial will continue to get their
old datasources.
LP: #1766401
|
|
In looking at some boot time for Xenial, Artful and Bionic, we noticed
some long amounts of time that appeared to be part of the DataSource but
we related to resolving URLs. In Artful and Bionic, there was an issue
(bug: #1739672) that resulted in slow getaddrinfo() calls when
systemd-resolved was in use. This patch adds two events that track time
for datasource.setup_datasource() and datasource.activate_datasource()
Additionally use log_time() to wrapper util.is_resolvable_url() which
leaves info in cloud-init.log about how much time was spent.
|
|
When images are deployed from template in a production environment
the artifacts of the provisioning stage (provisioningConfiguration.cfg)
that cloud-init referenced are cleaned up. However, when provisioned
in "debug" mode (internal to IBM) the artifacts are left.
This changes the 'is_ibm_provisioning' implementations in both
ds-identify and in the IBM datasource to identify the provisioning
stage more correctly. The change is to consider provisioning only
if the provisioing file existed and there was no log file or
the log file was older than this boot.
LP: #1767166
|
|
The cloud-init-local.service expects that any network device name changes
have already been completed by the kernel or udev daemon.
In some situations we've found that the renaming of interfaces from kernel
names (eth0, eth1, etc) to their persistent names (eno1, ens3, enp0s1,
etc) may happen after cloud-init-local has started where it reads values
from sysfs about what network devices are present, and which device to use
as a fallback nic.
Subsequently, cloud-init-local would write out network configuration for a
kernel device name which would no longer be present by the time that
networking services start to bring up the devices. The result is that the
instance does not get networking configured. Prior to use of
systemd-networkd, the Ubuntu 'networking.service' unit included a call to
udevadm settle which is why this race is not seen on a Xenial system.
This change adds the ability to detect if an interface has a stable name,
if if we find one without stable names and stable names have not been
disabled (net.ifnames=0 in /proc/cmdline), then cloud-init will invoke
udevadm settle.
LP: #1766287
|
|
This adds information to the IBMCloud datasource describing the
6 different scenarios that it is expected to handle.
|
|
BOOTPROTO=dhcp in sysconfig enables DHCPv4 and we should not do this
implicitly when 'dhcp6' subnet is specified. In case both dhcpv4 and
dhcpv6 are needed users should specify both:
subnets:
- type: dhcp6
- type: dhcp
Fix the current code and add a dhcpv6 only test.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
|
|
Ubuntu minimal images do not have iproute2, so correctly identify
our dependency on it.
LP: #1766711
|
|
cloud-init and mdata-get each have their own implementation of the
SmartOS metadata protocol. If cloud-init and other services that call
mdata-get are run concurrently, crosstalk on the serial port can cause
them both to become confused.
This change makes it so that cloud-init uses the same cooperative
locking scheme that's used by mdata-get, thus preventing cross-talk
between mdata-get and cloud-init.
For testing, a VM running on a SmartOS host and pyserial are required.
If the tests are run on a platform other than SmartOS, those that use a
real serial port are skipped. pyserial remains commented in
requirements.txt because most testers will not be running atop SmartOS.
LP: #1746605
|
|
There are three potential sources of the hostname, one of which is
documented SmartOS's vmadm(1M) via the hostname property. That
property's value is retrieved via the sdc:hostname key. The other
two sources for the hostname are a hostname key in customer_metadata
and the VM's uuid (sdc:uuid). Of these three, the sdc:hostname value
is not used in a meaningful way by DataSourceSmartOS.
This fix changes the fallback mechanism when hostname is not
specified in customer_metadata. The order of precedence for setting
the hostname is now 1) hostname in customer_metadata,
2) sdc:hostname, then 3) sdc:uuid.
LP: #1765085
|
|
If customer_metadata has no keys, the KEYS request returns an empty
string. Callers of the list() method expect a list to be returned and
will give a stack trace if this expectation is not met.
LP: #1763480
|
|
validate_cloudconfig_schema with strict=True would not actually validate
if there was no jsonschema available. That seems kind of strange.
The change here is to make it raise an exception if strict was passed in.
And then to fix the one test that needed a skipIfJsonSchema wrapper.
|
|
This admittedly does a fairly extensive re-factor to simply add a newline
to the end of sshd_config.
It makes the ssh_config updating portion of set_passwords more testable
and adds tests for that.
The new function is in 'update_ssh_config_lines' which allows you
to update a config with multiple changes even though only a single one
is currently used.
We also only restart the ssh daemon now if a change was made to the
config file. Before it was always restarted if the user specified
a value for ssh_pwauth other than 'unchanged'.
Thanks to Lorens Kockum for initial diagnosis and patch.
LP: #1677205
|
|
This enables warnings produced by pylint for unused variables (W0612),
and fixes the existing errors.
|
|
Just add some documentation to readthedocs for AliYun.
|
|
runcmd, bootcmd, snap/commands, ubuntu-advantage/commands would
log warning (and fail if strict) on duplicate values in the commands.
But those should be allowed. Example, it is perfectly valid to do:
runcmd: ['sleep 1', 'sleep 1']
LP: #1764264
|
|
The net-tools package is deprecated and will eventually be dropped. Use
"ip route", "link" or "address" instead of "ifconfig" or "route" calls.
Cloud-init can now run in an environment that no longer has net-tools.
This affects the network and route printing emitted to
cloud-config-output.log as well as the cc_disable_ec2_metadata module.
Additional changes:
- separate readResource and resourceLocation into standalone test
functions
- Fix ipv4 address rows to report scopes represented by ip addr show
- Formatted route/address ouput now handles multiple ipv4 and ipv6
addresses on a single interface
Co-authored-by: James Hogarth <james.hogarth@gmail.com>
Co-authored-by: Robert Schweikert <rjschwei@suse.com>
|
|
If the metadata service in the host is down while a guest that uses
DataSourceSmartOS is booting, the request from the guest falls into the
bit bucket. When the metadata service is eventually started, the guest
has no awareness of this and does not resend the request. This results in
cloud-init hanging forever with a guest reboot as the only recovery
option.
This fix updates the metadata protocol to implement the initialization
phase, just as is implemented by mdata-get and related utilities. The
initialization phase includes draining all pending data from the serial
port, writing an empty command and getting an expected error message in
reply. If the initialization phase times out, it is retried every five
seconds. Each timeout results in a warning message: "Timeout while
initializing metadata client. Is the host metadata service running?" By
default, warning messages are logged to the console, thus the reason for a
hung boot is readily apparent.
LP: #1667735
|
|
ext3 is not able to support file system sizes that are needed in Joyent's
cloud. For the default block size of 4k, the maximum filesystem size
for ext3 is 2^32 * 4096 = 16 TiB.
This changes the default file system type from ext3 to ext4.
LP: #1763511
|
|
Python has deprecated these invalid string literals now
https://bugs.python.org/issue27364
and pycodestyle is identifying them with a W605 warning.
https://github.com/PyCQA/pycodestyle/pull/676
So basically, any use of \ not followed by one of [\'"abfnrtv]
or \ooo (octal) \xhh (hex) or a newline is invalid. This is most
comomnly seen for us in regex. To solve, you either:
a.) use a raw string r'...'
b.) correctly escape the \ that was not intended to be interpreted.
|
|
In bash shells with bash_completion enabled, now the cloud-init
sub commands and parameters/flags will be shown.
|
|
This tool is used to assist during the creation of ubuntu packages for
release testing. Address the following on the command-line:
* --help option now print usage
* Add --orig-tarball which creates named output file
cloud-init_<release-version>.orig.tar.gz
* drop unused --verbose option
|
|
If a file passed to render_from_file had non-ascii text then
jinja in python2 would decode as ascii, which would cause
UnicodeDecodeError. This issue can be re-created in python2
with just:
'can\xe2\x80\x99t'.decode()
The solution here is to explicitly pass in unicode supporting
type (py3 str, py2 unicode). Those are six.text_type.
Then jinja does not try to decode.
The reason we hit this is that load_file calls decode_binary.
decode_binary believes it has no work to do if it got a six.string_types.
isinstance('can\xe2\x80\x99t', six.string_types) == True
So it returns the original string which will blow up for jinja.
Our fix here then is to load the file in binary mode and explicitly
decode it to utf-8. Then in python2 we'll have a unicode type
and in python3 we'll have a string type.
|
|
Add a base NTP client configuration dictionary and allow Distro
specific changes to be merged. Add a select client function which
implements logic to preferr installed clients over clients which
need to be installed. Also allow distributions to override the
cloud-init defaults.
LP: #1749722
|
|
When filing a bug with apport, this allows the user to choose
Brightbox, IBM, LXD, or OpenTelekomCloud as their cloud.
|
|
Fix integraiton test logic for ec2 to look for network and
availability-zone data under the key path
'ds'=>'meta-data' instead of just 'ds' when parsing instance-data.json.
|
|
Integration tests previously had a logic path that was unexercised on
jenkins because we were on an older version of lxc. With an upgrade to lxd
version 3.0 we need to bump pylxd dependency pin and fix a typo in
integration tests which checked the lxd version.
|
|
LP: #1420018
|
|
The zfs/zpool commands will hang for 10 seconds if /dev/zfs is not
present (bug 1760173). This is a common occurence for containers
using zfs as rootfs. Additionally handle missing zpool command or
other errors that may occur while executing the zpool command.
|
|
Fix link to external openstack resource and to internal vendor data.
LP: #1721660
|
|
Bump the version in cloudinit/version.py to be 18.2 and update ChangeLog.
LP: #1759318
|
|
This takes the same basic check that is in ds-identify. If the
DMI system manufacturer (aka sys_vendor) is not 'Hetzner', then exit
out of the datasource's get_data quickly.
|
|
This just correctly adds the missing dependency on isc-dhcp-client.
That package is used via 'dhclient' from cloudinit/net/dhcp.py.
LP: #1759307
|
|
Previously there was no support at all for zfs file system. With this
change it is now possible to use the resizefs module to grow a zpool to
its maximum partition size on FreeBSD.
LP: #1721243
|
|
Replace regressed cc_puppet functionality from a1f678f8.
The following content was inadvertently dropped:
- chown /var/lib/puppet/ssl as puppet:root.
- Automatic creation of /var/lib/puppet/ssl/certs
|
|
This just got missed in the IBMCloud datasource addition.
Add it to the builtin list of datasources.
|
|
This adds a specific IBM Cloud datasource.
IBM Cloud is identified by:
a.) running on xen
b.) one of a LABEL=METADATA disk or a LABEL=config-2 disk with
UUID=9796-932E
The datasource contains its own config-drive reader that reads
only the currently supported portion of config-drive needed for
ibm cloud.
During the provisioning boot, cloud-init is disabled.
See the docstring in DataSourceIBMCloud.py for more more information.
|
|
Ubuntu 16.04 (xenial) does not have jsonschema installed by default. As
it is listed in requirements, the tox environment will always have it
installed.
Add the helper tools/pipremove that removes pip packages. Then use that
to remove jsonschema without noise of always running and ignoring a
'pip uninstall jsonschema'.
|
|
The recently added snap and ubuntu_advantage modules had unit tests
that exercised jsonschema. Those throw error if jsonschema is
not present. Fix to skip in that scenario.
|
|
The recent change to exception_cb missed this caller.
The result was a slow test.
|
|
Reducing timeout to 1 second as IMDS responds within a handful
of milliseconds. Also get rid of max_retries to prevent exiting
out of polling loop early due to IMDS outage / upgrade.
Reduce Azure PreProvisioning HTTP timeouts during polling to
avoid waiting an extra minute.
LP: #1752977
|
|
In commit e9e8616, there was an inversion of the logic of the
exception_cb return value meaning, breaking the (network) OpenStack
DataSource, which implemented exception_cb as should_retry_cb, returning
True when a retry should be done and False when the retry loop should
be broken and the exception reraised again immediately.
The OpenStack DS was the only user of this callback at the time and not
touched by the commit (nor did the commit message mention an intended
change), so this almost certainly happened by mistake.
These days, we have a second user of the callback in DataSourceScaleway.
It uses the new logic, so it needs change if we fix the meaning of the
return value.
This patch reverts the meaning of url_helper.read_url() execption_cb
to the old semantics. It updates the comment and adjusts the Scaleway
datasource.
The patch has been tested on Open Telekom Cloud (which uses the
OpenStack network Datasource) where previously a missing user_data
and network_data.json would be retried 6 times each despite them
not being present (they are optional!) and the server repsonding
with a correct 404. After the patch, boot times are 10s faster,
as we no longer pointlessly retry these files.
LP: #1702160
LP: #1298921
|
|
ubuntu-advantage-tools is a package for enabling and disabling extended
support services such as Extended Security Maintenance (ESM), Canonical
Livepatch and FIPS certified PPAs. Simplify Ubuntu Advantage setup on
machines by allowing users to provide a list of ubuntu-advantage commands
in cloud-config.
|
|
In network config v1 format, there are dns values which are not bound to a
specific interface and do not map to the per-interface format in netplan.
To handle this case we render netplan configuration that duplicates the
DNS configuration on any interface that has a static network config. We
avoiding interfaces which have DHCP configuration which may provide
conflicting DNS values.
LP: #1750884
|
|
Open Telekom Cloud gen1 (Xen) hosts do not provide nova product
names in DMI but Xen HVM domU. They can however be safely identified
by the OpenTelekomCloud Chassis asset tag. OpenTelekomCloud does
use the network OpenStack DataSource, so we better detect it.
LP: #1756471
|
|
DataSource.get_hostname call signature changed to allow for metadata_only
parameter. The metadata_only=True parameter is passed to get_hostname
during init-local stage in order to set the system hostname if present in
metadata prior to initial network bring up.
Fix subclasses of DataSource which have overridden get_hostname to allow
for metadata_only param.
LP: #1757176
|
|
OpenNebulaNetwork.gen_conf() was previously returning ENI format.
This is updated to return netplan/v2 config.
The changes here also adds support for IPv6 configuration distributed
from OpenNebula and fixes some issues about nameserver information.
|
|
The Hetzner Cloud metadata service is an AWS-style service available
over HTTP via the link local address 169.254.169.254.
https://hetzner.com/cloud
https://docs.hetzner.cloud/
|
|
When 'ip=' or 'ip6=' is found on the kernel command line,
cloud-init will consider read network config from /run/net-*.conf files.
There are some iscsi-root scenarios where initramfs configures networking
but the ip= parameter is not present. 2 such cases are:
a.) static config in /etc/iscsi/iscsi.initramfs (copied into the
initramfs)
b.) iBft
This changes cloud-init to consider initramfs provided networking
information if:
* there are /run/net-* files and
* (ip= or ip6 is on the command line) or open-iscsi.interface file
exists.
LP: #1752391
|
|
|
|
|
