| Age | Commit message (Collapse) | Author |
|---|
|
RedHat dhcp client writes out rfc3442 classless-static-routes in a different format[1]
than what is found in isc-dhcp clients. This patch adds support for the RedHat format.
1. Background details on the format
https://bugzilla.redhat.com/show_bug.cgi?id=516325
https://github.com/vaijab/fedora-dhcp/blob/e83fb19c51765442d77fa60596bfdb2b3b9fbe2e/dhcp-rfc3442-classless-static-routes.patch#L252
https://github.com/heftig/NetworkManager/blob/f56c82d86122fc45304fc829b5f1e4766ed51589/src/dhcp-manager/nm-dhcp-client.c#L978
LP: #1850642
|
|
Sending a valid but empty v1 network config resulted in a
stacktrace during execution. Update the network_state
parse path to specific check if the 'config' key is None
(not present) versus being present but explicitly empty.
Also add some network_state unittests.
LP: #1852496
|
|
Router advertisements are required for the default route
to be set up, thus accept_ra should be enabled for
dhcpv6-stateful.
sysconf: IPV6_FORCE_ACCEPT_RA controls accept_ra sysctl.
eni: mode static and mode dhcp 'accept_ra' controls sysctl.
Add 'accept-ra: true|false' parameter to config v1 and
v2. When True: accept_ra is set to '1'. When False:
accept_ra is set to '0'. When not defined in config the
value is left to the operating system default.
This change also extend the IPv6 support to distinguish
between slaac and dhcpv6-stateless. SLAAC is autoconfig
without any options from DHCP, while stateless auto-configures
the address and the uses DHCP for other options.
LP: #1806014
LP: #1808647
|
|
Network v2 configuration for Azure will set both dhcp4 and
dhcp6 to False by default.
When IPv6 privateIpAddresses are present for an interface in Azure's
Instance Metadata Service (IMDS), set dhcp6: True and provide a
route-metric value that will match the corresponding dhcp4 route-metric.
The route-metric value will increase by 100 for each additional
interface present to ensure the primary interface has a route to IMDS.
Also fix dhcp route-metric rendering for eni and sysconfig distros.
LP: #1850308
|
|
Function return false for ipv6_dhcpv6-stateless|stateful,
the eni renderer does not add '6' to 'inet' which is
incorrect.
The subnet_is_ipv6() function is updated to also return
true if startswith('ipv6').
LP: #1848690
|
|
netplan introduced an 'info' subcommand which emits yaml describing
implemented features that indicate new or changed fields and values
in the yaml that it accepts. Previously, cloud-init emitted the key
'mtu6' for ipv6 MTU values. This is not correct and netplan will
fail to parse these values. Netplan as of 0.98 supports both the
info subcommand and the ipv6-mtu key.
This branch modifies the netplan renderer to collect the netplan
info output into a 'features' property which is a list of available
feature flags which the renderer can use to modify its output. If
the command is not available, no feature flags are set and
cloud-init will render IPv6 MTU values just as MTU for the subnet.
|
|
Here we replace uses of the pyyaml module directly with functions
provided by cloudinit.safeyaml. Also, change/move
cloudinit.util.yaml_dumps
to
cloudinit.safeyaml.dumps
LP: #1849640
|
|
In addition to ifup/ifdown the sysconfig renderer looks for
evidence that the sysconfig directory is properly populated.
This secondary check only considered RedHat specific location.
Fix this by adding a SUSE specific file and returning True if
either RedHat or SUSE file is present.
LP: #1849378
|
|
Commit e7b0e5f72 added support for configuring infiniband devices by
adding a new infiniband 'type'. This commit updates eni and sysconfig
renderers to consume this new type and configure infiniband devices
correctly.
LP: #1847114
|
|
|
|
Openstack subnets can be configured to use SLAAC by setting
ipv6_address_mode=dhcpv6-stateless. When this is the case
the sysconfig interface configuration should use
IPV6_AUTOCONF=yes and not set DHCPV6C=yes.
This change sets the subnets type property to the full
network['type'] from openstack metadata.
cloudinit/net/sysconfig.py and cloudinit/net/eni.py
are updated to support new subnet types:
- 'ipv6_dhcpv6-stateless' => IPV6_AUTOCONF=yes
- 'ipv6_dhcpv6-stateful' => DHCPV6C=yes
Type 'dhcp6' in sysconfig is kept for backward compatibility
with any implementations that set subnet_type == 'dhcp6'.
LP: #1847517
|
|
The change that introduced this issue was handling interfaces that are
bonded in the kernel, in a way that doesn't present as "a bond" to
userspace in the normal way. Both members of this "bond" will share a
MAC address, so we filter one of them out to avoid incorrect MAC address
collision warnings.
Unfortunately, the matching condition was too broad, so that change also
affected normal bonds and bridges. This change specifically excludes
bonds and bridges from that determination, to address that regression.
LP: #1846535
|
|
If an OS image provided an /etc/resolv.conf file that was not empty
cloud-init would read and re-write it with a cloud-init header even
if no DNS network configuration was provided (e.g. DHCP only).
This can cause problems for some network services which don't
ignore cloud-init's header.
LP: #1843634
|
|
The sysconfig renderer used the distro name directly which mean
some variants of distros were not considered supported. Fix this
by using util.system_info()['variant'] instead. Fix the list of
KNOWN_DISTROS value for redhat -> rhel.
LP: #1843584
|
|
Some network devices are transformed into a bond via kernel magic
and do not have the 'bonding' sysfs attribute, but like a bond they
have a duplicate MAC of other bond members. On Azure Advanced
Networking SRIOV devices are auto bonded and will have the same MAC
as the HyperV nic. We can detect this via the 'master' sysfs attribute
in the device sysfs path and this patch adds this to the list of devices
we ignore when enumerating device lists.
LP: #1844191
|
|
Add support for detecting netfailover[1] device 3-tuple in networking
layer. In the Oracle datasource ensure that if a provided network
config, either fallback or provided config includes a netfailover master
to remove any MAC address value as this can break under 3-netdev
as the other two devices have the same MAC.
1. https://www.kernel.org/doc/html/latest/networking/net_failover.html
|
|
This refactors read_initramfs_config to support multiple different types
of initramfs network configuration. It introduces an
InitramfsNetworkConfigSource abstract base class. There is currently a
single sub-class, KlibcNetworkConfigSource, which contains the logic
which previously was directly within read_initramfs_config.
|
|
The function generate_fallback_config is used by Azure by default when
not consuming IMDS configuration data. This function is also used by any
datasource which does not implement it's own network config. This simple
fallback configuration sets up dhcp on the most likely NIC. It will now
emit network v2 instead of network v1.
This is a step toward moving all components talking in v2 and allows us
to avoid costly conversions between v1 and v2 for newer distributions
which rely on netplan.
|
|
Previously "cmdline" network configuration could be either
user-specified network-config=... configuration data, or
initramfs-provided configuration data. Before data sources could modify
the order in which network config sources were considered, this
conflation didn't matter (and, indeed, in the default data source
configuration it will continue to not matter).
However, it _is_ desirable for a data source to be able to specify that
its network configuration should be preferred over the
initramfs-provided network configuration but still allow explicit
network-config=... configuration passed to the kernel cmdline to
continue to override both of those sources.
(This also modifies the Oracle data source to use read_initramfs_config
directly, which is effectively what it was using
read_kernel_cmdline_config for previously.)
|
|
If input to network_state.parse_net_config_data was netplan (v2 yaml)
then the network state would lose the mtu information on bond or vlan.
LP: #1836949
|
|
On systems with many interfaces, processing udev events may take a while.
Cloud-init expects devices included in a provided network-configuration
to be present when attempting to configure them. This patch adds a step
in net configuration where it will check for devices provided in the
configuration and if not found, issue udevadm settle commands to wait
for them to appear.
Additionally, the default path for udev persistent network rules
70-persistent-net.rules may also be written to systems which include
the 75-net-generator.rules. During boot, cloud-init and the
generator may race and interleave values causing issues. OpenSUSE
will now use a newer file, 85-persistent-net-cloud-init.rules which
will take precedence over values created by 75-net-generator and
avoid collisions on the same file.
LP: #1817368
|
|
The EphemeralDHCP context manager did not parse or handle
rfc3442 classless static routes which prevented reading
datasource metadata in some clouds. This branch adds support
for extracting the field from the leases output, parsing the
format and then adding the required iproute2 ip commands to
apply (and teardown) the static routes.
LP: #1821102
|
|
bonds may inherit mac address from a physical interface
LP: #1812857
|
|
Currently, only a few bonding parameters can be configured on
sysconfig systems. This patch aims to support more parameters
documented on the docs site.
|
|
Previous versions of netplan included a misspelling for the
bond parameter around gratuitous-arp. This has been fixed and released
and cloud-init needs to accept both values. This branch fixes the
key that will be rendered and transforms the previous misspelling
when capturing network_state.
LP: #1827238
|
|
Currently on 18.04, running tox -e py27 will spew errors like:
.tests/unittests/test_net.py:2649: YAMLLoadWarning: calling yaml.load()
without Loader=... is deprecated, as the default Loader is unsafe.
Please read https://msg.pyyaml.org/load for full details.
The change here just uses cloud-init's yaml, which does safeloading
by default.
|
|
Restrict the sysconfig renderer availabily to known distros.
Ubuntu/Debian systems may include network-manager but they do
not have support for reading sysconfig network output; that is
enabled via a Network-Manager plugin: ifcfg-rh which is not
available in Ubuntu/Debian.
LP: #1819994
|
|
For writing IPv6 addresses to ifcfg-* the name "IPV6ADDR" is used. For
secondary IPs the value for "IPV6ADDR_SECONDARIES" is set. On SUSE based
distributions the names "IPADDR6" and "IPADDR6_$SOMELABEL" need to be
used.
|
|
When rendering ipv6 static routes in eni format the
post-up/pre down commands were not correct for ipv6.
LP: #1818669
|
|
When the network configuration has a default route configured and
another network device that is configured with dhcp, SUSE sysconfig
output should not accept the default route provided by the dhcp
server.
LP: #1812117
|
|
When providing netplan configuration to cloud-init, the internal
network state would enable DHCP if the 'dhcp' key was present in
the source config. In netplan, dhcp[46] is a boolean and the
value of the boolean should control whether DHCP is enabled rather
than the presence of the key. This issue leaded to inconsistant
sysconfig/network-scripts on fedora. 'BOOTPROTO' was always 'dhcp',
even if the address config was static.
After this change a dhcp subnet is added only if the 'dhcp' setting
in source cfg dict is True.
LP: #1818032
|
|
In test_ds_identify, don't mutate otherwise-static test data. When
running tests in a random order, this was causing failures due to
breaking preconditions for other tests.
In tests/helpers, reset logging level in tearDown. Some of the CLI
tests set the level of the root logger in a way that isn't correctly
reset.
For test_poll_imds_re_dhcp_on_timeout and
test_dhcp_discovery_run_in_sandbox_warns_invalid_pid, mock out
time.sleep; this saves ~11 seconds (or ~40% of previous test time!).
|
|
Cloud-init rendered netplan with duplicate aliases if a network config
included "global" nameserver/search values. Netplan uses can read yaml
files which do use aliaes but cloud-init did not render a single yaml
dictionary, instead it combined yaml sections into a single document
which sometimes resulted in duplicate aliases being present.
This branch introduces a yaml SafeDumper class which can set the
'ignore_aliases' attribute. This is not enabled by default but callers
to util.yaml_dumps can pass a boolean to toggle this. The netplan
render uses noalias=True and the resulting yaml output does not contain
any aliases.
LP: #1815051
|
|
ONBOOT is not recognized on openSUSE and SUSE Linux Enterprise,
add the STARTMODE setting
LP: #1799540
|
|
Writing the file with no dns information may prevent distro tools
from writing a resolv.conf file with dns information obtained from
a dhcp server.
|
|
The 'sysconfig' renderer is activated if, and only if, there's ifup and
ifdown commands present in its search dictonary or the network-scripts
configuration files are found. This patch adds a check for Network-
Manager configuration file as well.
This solution is based on the use of the plugin 'ifcfg-rh' present in
Network-Manager and is designed to support Fedora 29 or other
distributions that also replaced network-scripts by Network-Manager.
|
|
cloud-init uses dhclient to fetch the DHCP lease so it can extract
DHCP options. dhclient creates the leasefile, then writes to it;
simply waiting for the leasefile to appear creates a race between
dhclient and cloud-init. Instead, wait for dhclient to be parented by
init. At that point, we know it has written to the leasefile, so it's
safe to copy the file and kill the process.
cloud-init creates a temporary directory in which to execute dhclient,
and deletes that directory after it has killed the process. If
cloud-init abandons waiting for dhclient to daemonize, it will still
attempt to delete the temporary directory, but will not report an
exception should that attempt fail.
LP: #1794399
|
|
It is possible to have a metric value in a per-subnet route.
This is currently missing in all renderers. Update each
renderer to emit the correct metric value from the config.
LP: #1805871
|
|
We add a new Optional parameter: connectivity_url
This is used in __enter__ to verify if a connection already exists.
If it does exist, no operations are performed.
|
|
In commitish 9073951 azure datasource tried to leverage stale DHCP
information obtained from EphemeralDHCPv4 context manager to report
updated provisioning status to the fabric earlier in the boot process.
Unfortunately the stale ephemeral network configuration had already been
torn down in preparation to bring up IMDS network config so the report
attempt failed on timeout.
This branch introduces obtain_lease and clean_network public methods on
EphemeralDHCPv4 to allow for setup and teardown of ephemeral network
configuration without using a context manager. Azure datasource now uses
this to persist ephemeral network configuration across multiple contexts
during provisioning to avoid multiple DHCP roundtrips.
|
|
Previously we explicitly excluded mac address '00:00:00:00:00:00'.
But then some nics (tunl0 and sit0) ended up having a mac address like
'00:00:00:00'.
The change here just ignores all 00[:00[:00...]].
LP: #1796917
|
|
OpenStack ironic references Infiniband interfaces via a 6 byte 'MAC
address' formed from bytes 13-15 and 18-20 of interface's hardware
address. This address is used as the ethernet_mac_address of Infiniband
links in network_data.json in configdrives generated by OpenStack nova.
We can use this address to map links in network_data.json to their
corresponding interface names.
When generating interface configuration files, we need to use the
interface's full hardware address as the HWADDR, rather than the 6 byte
MAC address provided by network_data.json.
This change allows IB interfaces to be referenced in this dual mode - by
MAC address and hardware address, depending on the context.
Support TYPE=InfiniBand for sysconfig configuration of IB interfaces.
|
|
On OpenStack based OVH public cloud, we got DHCP response with
fixed-address 54.36.113.86;
option subnet-mask 255.255.255.255;
option routers 54.36.112.1;
The router clearly is not on the subnet. So 'ip' would fail when
we tried to add the default route.
The solution here is to add an explicit route on that interface
to the router and then add the default route.
Also add 'bgpovs' to the list of 'physical' types for OpenStack
network configuration. That type is used on OVH public cloud.
LP: #1792415
|
|
In many cases, cloud-init uses 'util.subp' to run a subprocess.
This is not really desirable in our unit tests as it makes the tests
dependent upon existance of those utilities.
The change here is to modify the base test case class (CiTestCase) to
raise exception any time subp is called. Then, fix all callers.
For cases where subp is necessary or actually desired, we can use it
via
a.) context hander CiTestCase.allow_subp(value)
b.) class level self.allowed_subp = value
Both cases the value is a list of acceptable executable names that
will be called (essentially argv[0]).
Some cleanups in AltCloud were done as the code was being updated.
|
|
Multiple distros use sysconfig format but have different content
and paths to certain files. Update distros to specify these
template paths in their renderer_configs dictionary.
|
|
When converting network config v1 to netplan, we were not correctly
rendering the 'macaddress' key on a bond. Not that the difference
in spelling between v1 'mac_address' and v2 'macaddress' is intentional.
Also fixed here is rendering of the macaddress for bridges.
LP: #1784699
|
|
When network configuration for any interface defines maximum transmission
values (MTU) the netplan, eni and sysconfig renders will take into account
any device-level, or subnet-level mtu values.
When network configuration has conflicting device-level and ipv4 subnet
mtu values, the subnet-specific value is honored and a warning will be
logged about any ignored device-level setting.
LP: #1774666
|
|
When attempting to apply network configuration for SmartOS's container
platform, cloud-init would not identify nics. The nics on provided
in this container service do not have 'addr_assign_type'. That
was being interpreted as being a "stolen" mac, and would be filtered
out by get_interfaces.
|
|
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
|
|
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>
|
