Problem description:
--------------------
On Ubuntu, state 'network.managed' doesn't add slaves
to bonding master interface after it is UP
Example pillars:
linux_interface_bond0:
network.managed:
- enabled: True
- name: bond0
- type: bond
- mode: active-backup
- slaves: enp8s0f1
- require:
- network: linux_interface_enp8s0f1
linux_interface_enp8s0f1:
network.managed:
- enabled: True
- name: enp8s0f1
- type: slave
- master: bond0
Example result:
3: enp8s0f1: <BROADCAST,MULTICAST> mtu 1500 qdisc mq state DOWN mode DEFAULT group default qlen 1000
link/ether 0c:c4:7a:a8:d0:37 brd ff:ff:ff:ff:ff:ff
42: bond0: <NO-CARRIER,BROADCAST,MULTICAST,MASTER,UP> mtu 9000 qdisc noqueue master ovs-system state DOWN mode DEFAULT group default qlen 1000
link/ether ee:ae:74:0d:17:29 brd ff:ff:ff:ff:ff:ff
...
Bonding Mode: fault-tolerance (active-backup)
Primary Slave: None
Currently Active Slave: None
...
Workarounds:
------------
For Ubuntu, it is suggested to restart 'networking' service
to make changes take effect.
Another common workaround is to reboot the machine.
These workarounds are not suitable in some cases, when only non-primary
interfaces are added to bonding interface, so the connectivity to
the salt minion is not broken during state execution.
Proposed solution:
------------------
Let's use 'ifenslave' tool to add slave interfaces to the bonding
interface right after the bonding interface is created/restarted.
Add two new missing parameters to s3 fileclient (https_enable and path_style). They allow to use s3 compatible object storages with path style and http protocol.
RedHat updated cffi for RHEL/CentOS 7.3, which causes a pair of warnings
on the pygit2 import. This warning is spit out to the CLI on several
commands, when they result in the git fileserver backend to be loaded.
This commit squelches that warning as it is just noise. Upgrading
pygit2/libgit2 would solve this, but that is not likely to happen in
RHEL/CentOS 7.
