valitydev/salt
14
0
Fork 0
mirror of https://github.com/valitydev/salt.git synced 2024-11-09 01:36:48 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
b1bd992b1e
salt/doc/topics/releases/oxygen.rst
Bo Maryniuk b1bd992b1e Add more doc fixes
2018-01-15 14:31:25 +01:00

1510 lines
46 KiB
ReStructuredText
Raw Blame History

:orphan:
====================================
Salt Release Notes - Codename Oxygen
====================================
Lots of Docker Improvements
---------------------------
Much Improved Support for Docker Networking
===========================================
The :py:func:`docker_network.present <salt.states.docker_network.present>`
state has undergone a full rewrite, which includes the following improvements:
Full API Support for Network Management
---------------------------------------
The improvements made to input handling in the
:py:func:`docker_container.running <salt.states.docker_container.running>`
state for 2017.7.0 have now been expanded to :py:func:`docker_network.present
<salt.states.docker_network.present>`. This brings with it full support for all
tunable configuration arguments.
Custom Subnets
--------------
Custom subnets can now be configured. Both IPv4 and mixed IPv4/IPv6 networks
are supported. See :ref:`here <salt-states-docker-network-present-ipam>` for
more information.
Network Configuration in :py:func:`docker_container.running` States
-------------------------------------------------------------------
A long-requested feature has finally been added! It is now possible to
configure static IPv4/IPv6 addresses, as well as links and labels. See
:ref:`here <salt-states-docker-container-network-management>` for more
information.
.. note::
While the ``containers`` argument to :py:func:`docker_network.present`
will continue to be supported, it will no longer be the recommended way of
ensuring that a container is attached to a network.
Improved Handling of Images from Custom Registries
==================================================
Rather than attempting to parse the tag from the passed image name, Salt will
now resolve that tag down to an image ID and use that ID instead.
.. important::
Due to this change, there are some backward-incompatible changes to image
management. See below for a full list of these changes.
Backward-incompatible Changes to Docker Image Management
********************************************************
Passing image names to the following functions must now be done using separate
``repository`` and ``tag`` arguments:
- :py:func:`docker.build <salt.modules.dockermod.build>`
- :py:func:`docker.commit <salt.modules.dockermod.commit>`
- :py:func:`docker.import <salt.modules.dockermod.import_>`
- :py:func:`docker.load <salt.modules.dockermod.load>`
- :py:func:`docker.tag <salt.modules.dockermod.tag_>`
- :py:func:`docker.sls_build <salt.modules.dockermod.sls_build>`
Additionally, the ``tag`` argument must now be explicitly passed to the
:py:func:`docker_image.present <salt.states.docker_image.present>` state,
unless the image is being pulled from a docker registry.
State and Execution Module Support for ``docker run`` Functionality
===================================================================
The :py:func:`docker_container.running <salt.states.docker_container.running>`
state is good for containers which run services, but it is not as useful for
cases in which the container only needs to run once. The ``start`` argument to
:py:func:`docker_container.running <salt.states.docker_container.running>` can
be set to ``False`` to prevent the container from being started again on a
subsequent run, but for many use cases this is not sufficient. Therefore, the
:py:func:`docker.run_container <salt.modules.dockermod.run_container>`
remote-execution function was added. When used on the Salt CLI, it will return
information about the container, such as its name, ID, exit code, and any
output it produces.
State support has also been added via the :py:func:`docker_container.run
<salt.states.docker_container.run>` state. This state is modeled after the
:py:func:`cmd.run <salt.states.cmd.run>` state, and includes arguments like
``onlyif``, ``unless``, and ``creates`` to control whether or not the container
is run.
Full API Support for :py:func:`docker.logs <salt.modules.dockermod.logs>`
=========================================================================
This function now supports all of the functions that its Docker API counterpart
does, allowing you to do things like include timestamps, and also suppress
stdout/stderr, etc. in the return.
`start` Argument Added to :py:func:`docker.create <salt.modules.dockermod.create>` Function
===========================================================================================
This removes the need to run :py:func:`docker.start
<salt.modules.dockermod.start_>` separately when creating containers on the
Salt CLI.
.. code-block:: bash
salt myminion docker.create image=foo/bar:baz command=/path/to/command start=True
Use SaltSSH Minions like regular Master-Minions
-----------------------------------------------
The Master process can now also call SSH minions as if they were connected to
the master using ZeroMQ. By setting `enable_ssh_minions: True` in the the
master config file, the master will create a SaltSSH client process which
connects to the minion and returns the output for the `salt` commandline to use
like a regular minion. This can be used anywhere the LocalClient is used.
Comparison Operators in Package Installation
--------------------------------------------
Salt now supports using comparison operators (e.g. ``>=1.2.3``) when installing
packages on minions which use :mod:`yum/dnf <salt.modules.yumpkg>` or
:mod:`apt <salt.modules.aptpkg>`. This is supported both in the
:py:func:`pkg.installed <salt.states.pkg.installed>` state and in the ``pkg.install``
remote execution function.
:ref:`Master Tops <master-tops-system>` Changes
-----------------------------------------------
When both :ref:`Master Tops <master-tops-system>` and a
:ref:`Top File <states-top>` produce SLS matches for a given minion, the matches
were being merged in an unpredictable manner which did not preserve ordering. This has
been changed. The top file matches now execute in the expected order, followed
by any master tops matches that are not matched via a top file.
To make master tops matches execute first, followed by top file matches, set
the new :conf_minion:`master_tops_first` minion config option to ``True``.
Several Jinja Filters Renamed
-----------------------------
The following Jinja filters (originally added in 2017.7.0) have been renamed
due to the fact that they were inaccurately named when initially added. The
original names will be supported until the Neon release of Salt.
- :jinja_ref:`rand_str` renamed to :jinja_ref:`random_hash`
- :jinja_ref:`jinja_decode_dict` renamed to :jinja_ref:`jinja_encode_dict`
- :jinja_ref:`jinja_decode_list` renamed to :jinja_ref:`jinja_encode_list`
Return Codes for Runner/Wheel Functions
---------------------------------------
When using :ref:`orchestration <orchestrate-runner>`, runner and wheel
functions used to report a ``True`` result if the function ran to completion
without raising an exception. It is now possible to set a return code in the
``__context__`` dictionary, allowing runner and wheel functions to report that
they failed. Here's some example pseudocode:
.. code-block:: python
def myrunner():
...
do stuff
...
if some_error_condition:
__context__['retcode'] = 1
return result
Variable Update Intervals for Fileserver Backends
-------------------------------------------------
Prior to this release, fileservers would be updated as part of a dedicated
"maintenance" process, in which various routine maintenance tasks were
performed. This tied the update interval to the :conf_master:`loop_interval`
config option, and also forced all fileservers to update at the same interval.
Oxygen adds the following configuration options for the various fileserver
backends:
- :conf_master:`roots_update_interval`
- :conf_master:`azurefs_update_interval`
- :conf_master:`gitfs_update_interval`
- :conf_master:`hgfs_update_interval`
- :conf_master:`minionfs_update_interval`
- :conf_master:`s3fs_update_interval`
- :conf_master:`svnfs_update_interval`
These allow for update intervals to be set for each individual backend. The
default value for each of these is 60 seconds.
In addition, for :ref:`GitFS <tutorial-gitfs>` it is also possible to apply
intervals to individual remotes. See :ref:`here <gitfs-update-intervals>` for
examples.
.. note::
git_pillar does not yet support variable update intervals, this is targeted
for the next feature release (Fluorine).
LDAP via External Authentication Changes
----------------------------------------
In this release of Salt, if LDAP Bind Credentials are supplied, then
these credentials will be used for all LDAP access except the first
authentication when a job is submitted. The first authentication will
use the user's credentials as passed on the CLI. This behavior is to
accommodate certain two-factor authentication schemes where the authentication
token can only be used once.
In previous releases the bind credentials would only be used to determine
the LDAP user's existence and group membership. The user's LDAP credentials
were used from then on.
Stormpath External Authentication Removed
-----------------------------------------
Per Stormpath's announcement, their API will be shutting down on 8/17/2017 at
noon PST so the Stormpath external authentication module has been removed.
https://stormpath.com/oktaplusstormpath
New (Proxy) Minion Configuration Options
----------------------------------------
To be able to connect the Minion to the Master using a certain source IP address
or port, the following options have been added:
- :conf_minion:`source_interface_name`
- :conf_minion:`source_address`
- :conf_minion:`source_ret_port`
- :conf_minion:`source_publish_port`
:conf_minion:`environment` config option renamed to :conf_minion:`saltenv`
--------------------------------------------------------------------------
The :conf_minion:`environment` config option predates referring to a salt
fileserver environment as a **saltenv**. To pin a minion to a single
environment for running states, one would use :conf_minion:`environment`, but
overriding that environment would be done with the ``saltenv`` argument. For
consistency, :conf_minion:`environment` is now simply referred to as
:conf_minion:`saltenv`. There are no plans to deprecate or remove
:conf_minion:`environment`, if used it will log a warning and its value will be
used as :conf_minion:`saltenv`.
:conf_minion:`lock_saltenv` config option added
-----------------------------------------------
If set to ``True``, this option will prevent a minion from allowing the
``saltenv`` argument to override the value set in :conf_minion:`saltenv` when
running states.
Failed Minions for State/Function Orchestration Jobs Added to Changes Dictionary
--------------------------------------------------------------------------------
For orchestration jobs which run states (or run remote execution functions and
also use a :ref:`fail function <orchestrate-runner-fail-functions>` to indicate
success or failure), minions which have ``False`` results were previously
included as a formatted string in the comment field of the return for that
orchestration job. This made the failed returns difficult to :ref:`parse
programatically <orchestrate-runner-parsing-results-programatically>`. The
failed returns in these cases are now included in the changes dictionary,
making for much easier parsing.
New Grains
----------
New core grains have been added to expose any storage inititator setting.
The new grains added are:
* ``fc_wwn``: Show all fibre channel world wide port names for a host
* ``iscsi_iqn``: Show the iSCSI IQN name for a host
* ``swap_total``: Show the configured swap_total for Linux, \*BSD, OS X and Solaris/SunOS
Salt Minion Auto-discovery
------------------------
Salt Minion now no longer need to be configured against a specific DNS name or IP address of a Master.
For this feature Salt Master now requires port 4520 for UDP broadcast packets to be opened
and the Salt Minion be able to send UDP packets to the same port.
Configuration
=============
By default, automatic discovery is disabled.
..warning::
Due to the current limitations that will be changing in a future, before you turn on auto-discovery,
make sure your network is secured and trusted.
Auto-discovery is configured on Master and Minion. Both of them are configured via ``discovery`` option
as follows:
**Master configuration**
To simply enable it by default, accepting all kinds of minions, it is enough to set it to True:
.. code-block:: yaml
discovery: true
Sub-option `mapping` allows auto-discovery to help Minions to find a proper Master for them. The mapping contains
just an arbitrary set of key/value strings that would match either any of them or all of them. Default is empty.
Example:
.. code-block:: yaml
discovery:
mapping:
description: SES 5.0
node: 1
It is also possible to change the port used from the default of ``4520``, by
setting a ``port`` option under the Master's ``discovery`` configuration:
.. code-block:: yaml
discovery:
port: 4567
.. note::
When using a port number other than the default, the Minion's ``discovery``
configuraton must *also* have a port specified, otherwise the Minion will
still attempt to contact the Master on port ``4520``.
**Minion configuration**
In addition to the ``mapping`` and ``port`` options, the following additional options are available to Minions:
- ``attempts`` will set how many "ping" broadcast requests should be sent to the network, waiting for any Master
response. Each attempt takes a couple of seconds, so raising it to the higher value will
result to a slower startup of the Minion. Generally, on a properly configured network, the autodiscovery is
supposed to be all set right after the first attempt. Default is set to 3 attempts.
- ``match`` can have values ``all`` or ``any``. Match is always coming together with ``mapping``, and essentially means
"match any/all of mapping values to the target". Default is set to ``any``.
- ``pause`` is an interval in seconds between attempts. Default is set to five (5) seconds.
Connection to a type instead of DNS
===================================
By now each Minion was connecting to a Master by DNS or IP address. From now on it is possible
also to connect to a _type_ of a Master. For example, in a network there are three different
Masters, each corresponds for a particular niche or environment or specific role etc. The Minion
is supposed to connect only to one of those Masters that is described approriately.
To achieve such an effect, each `/etc/salt/master` configuration should have a `discovery` option,
which should have a `mapping` element with arbitrary key/value pairs. The same configuration should
be on the Minion, so then when mapping matches, Minion recognises Master as its connection target.
Example for Master configuration (`/etc/salt/master`):
.. code-block:: yaml
discovery:
mapping:
description: SES 5.0
node: 1
The example above describes a system that is running a particular product, where `description` is
an arbitrary key and `SES 5.0` is just a string. In order to match exactly this Master, the
following configuration at Minion should be present:
.. code-block:: yaml
discovery:
match: all # Can be "all" or "any"
mapping:
description: SES 5.0
node: 1
Notice `match` criteria is set to `all`. This would mean that from all found Masters select only
that, which `description` is set to `SES 5.0` _and_ `node` is set to `1`. All other Masters will
be ignored.
Limitations
===========
This feature has a couple of _temporary_ limitations that are subject to change in the future:
- Only one Master on the network is supported. Currently the Minion cannot select which Master
out of few the same to choose. This will change to choosing the Master that is least loaded.
- Minions will accept _any_ master that matches connection criteria without any particular
security applied (priv/pub key check, signature, fingerprint etc). That implies that administrator
is expected to know his network and make sure it is clean.
Grains Changes
--------------
* The ``virtual`` grain identifies reports KVM and VMM hypervisors when running
an OpenBSD guest
New Modules
-----------
- :mod:`salt.modules.purefa <salt.modules.purefa>`
New NaCl Renderer
-----------------
A new renderer has been added for encrypted data.
New support for Cisco UCS Chassis
---------------------------------
The salt proxy minion now allows for control of Cisco USC chassis. See
the ``cimc`` modules for details.
New support for Cassandra v3
----------------------------
The ``cassandra_cql`` module now supports Cassandra v3 which has changed
its internal schema to define keyspaces and columns.
New salt-ssh roster
-------------------
A new roster has been added that allows users to pull in a list of hosts
for salt-ssh targeting from a ``~/.ssh`` configuration. For full details,
please see the ``sshconfig`` roster.
New GitFS Features
------------------
Two new features which affect how GitFS maps branches/tags to fileserver
environments (i.e. ``saltenvs``) have been added:
1. It is now possible to completely turn off Salt's default mapping logic
(aside from the mapping of the ``base`` saltenv). This can be triggered
using the new :conf_master:`gitfs_disable_saltenv_mapping` config option.
.. note::
When this is disabled, only the ``base`` saltenv and any configured
using :ref:`per-saltenv configuration parameters
<gitfs-per-saltenv-config>` will be available.
2. The types of refs which Salt will use as saltenvs can now be controlled. In
previous releases, branches and tags were both mapped as environments, and
individual commit SHAs could be specified as saltenvs in states (and when
caching files using :py:func:`cp.cache_file <salt.modules.cp.cache_file>`).
Using the new :conf_master:`gitfs_ref_types` config option, the types of
refs which are used as saltenvs can be restricted. This makes it possible to
ignore all tags and use branches only, and also to keep SHAs from being made
available as saltenvs.
Additional output modes
-----------------------
The ``state_output`` parameter now supports ``full_id``, ``changes_id`` and ``terse_id``.
Just like ``mixed_id``, these use the state ID as name in the highstate output.
For more information on these output modes, see the docs for the :mod:`Highstate Outputter <salt.output.highstate>`.
Windows Installer: Changes to existing config handling
------------------------------------------------------
Behavior with existing configuration has changed. With previous installers the
existing config was used and the master and minion id could be modified via the
installer. It was problematic in that it didn't account for configuration that
may be defined in the ``minion.d`` directory. This change gives you the option
via a checkbox to either use the existing config with out changes or the default
config using values you pass to the installer. If you choose to use the existing
config then no changes are made. If not, the existing config is deleted, to
include the ``minion.d`` directory, and the default config is used. A
command-line switch (``/use-existing-config``) has also been added to control
this behavior.
Windows Installer: Multi-master configuration
---------------------------------------------
The installer now has the ability to apply a multi-master configuration either
from the gui or the command line. The ``master`` field in the gui can accept
either a single master or a comma-separated list of masters. The command-line
switch (``/master=``) can accept the same.
Windows Installer: Command-line help
------------------------------------
The Windows installer will now display command-line help when a help switch
(``/?``) is passed.
Salt Cloud Features
-------------------
Pre-Flight Commands
===================
Support has been added for specified "preflight commands" to run on a VM before
the deploy script is run. These must be defined as a list in a cloud configuration
file. For example:
.. code-block:: yaml
my-cloud-profile:
provider: linode-config
image: Ubuntu 16.04 LTS
size: Linode 2048
preflight_cmds:
- whoami
- echo 'hello world!'
These commands will run in sequence **before** the bootstrap script is executed.
New salt-cloud Grains
=====================
When salt cloud creates a new minon, it will now add grain information
to the minion configuration file, identifying the resources originally used
to create it.
The generated grain information will appear similar to:
.. code-block:: yaml
grains:
salt-cloud:
driver: ec2
provider: my_ec2:ec2
profile: ec2-web
The generation of salt-cloud grains can be surpressed by the
option ``enable_cloud_grains: 'False'`` in the cloud configuration file.
Upgraded Saltify Driver
=======================
The salt-cloud Saltify driver is used to provision machines which
are not controlled by a dedicated cloud supervisor (such as typical hardware
machines) by pushing a salt-bootstrap command to them and accepting them on
the salt master. Creation of a node has been its only function and no other
salt-cloud commands were implemented.
With this upgrade, it can use the salt-api to provide advanced control,
such as rebooting a machine, querying it along with conventional cloud minions,
and, ultimately, disconnecting it from its master.
After disconnection from ("destroying" on) one master, a machine can be
re-purposed by connecting to ("creating" on) a subsequent master.
New Vagrant Driver
==================
The salt-cloud Vagrant driver brings virtual machines running in a limited
environment, such as a programmer's workstation, under salt-cloud control.
This can be useful for experimentation, instruction, or testing salt configurations.
Using salt-api on the master, and a salt-minion running on the host computer,
the Vagrant driver can create (``vagrant up``), restart (``vagrant reload``),
and destroy (``vagrant destroy``) VMs, as controlled by salt-cloud profiles
which designate a ``Vagrantfile`` on the host machine.
The master can be a very limited machine, such as a Raspberry Pi, or a small
VagrantBox VM.
New pillar/master_tops module called saltclass
----------------------------------------------
This module clones the behaviour of reclass (http://reclass.pantsfullofunix.net/), without the need of an external app, and add several features to improve flexibility.
Saltclass lets you define your nodes from simple ``yaml`` files (``.yml``) through hierarchical class inheritance with the possibility to override pillars down the tree.
**Features**
- Define your nodes through hierarchical class inheritance
- Reuse your reclass datas with minimal modifications
- applications => states
- parameters => pillars
- Use Jinja templating in your yaml definitions
- Access to the following Salt objects in Jinja
- ``__opts__``
- ``__salt__``
- ``__grains__``
- ``__pillars__``
- ``minion_id``
- Chose how to merge or override your lists using ^ character (see examples)
- Expand variables ${} with possibility to escape them if needed \${} (see examples)
- Ignores missing node/class and will simply return empty without breaking the pillar module completely - will be logged
An example subset of datas is available here: http://git.mauras.ch/salt/saltclass/src/master/examples
========================== ===========
Terms usable in yaml files Description
========================== ===========
classes A list of classes that will be processed in order
states A list of states that will be returned by master_tops function
pillars A yaml dictionnary that will be returned by the ext_pillar function
environment Node saltenv that will be used by master_tops
========================== ===========
A class consists of:
- zero or more parent classes
- zero or more states
- any number of pillars
A child class can override pillars from a parent class.
A node definition is a class in itself with an added ``environment`` parameter for ``saltenv`` definition.
**class names**
Class names mimic salt way of defining states and pillar files.
This means that ``default.users`` class name will correspond to one of these:
- ``<saltclass_path>/classes/default/users.yml``
- ``<saltclass_path>/classes/default/users/init.yml``
**Saltclass tree**
A saltclass tree would look like this:
.. code-block:: text
<saltclass_path>
├── classes
│ ├── app
│ │ ├── borgbackup.yml
│ │ └── ssh
│ │ └── server.yml
│ ├── default
│ │ ├── init.yml
│ │ ├── motd.yml
│ │ └── users.yml
│ ├── roles
│ │ ├── app.yml
│ │ └── nginx
│ │ ├── init.yml
│ │ └── server.yml
│ └── subsidiaries
│ ├── gnv.yml
│ ├── qls.yml
│ └── zrh.yml
└── nodes
├── geneva
│ └── gnv.node1.yml
├── lausanne
│ ├── qls.node1.yml
│ └── qls.node2.yml
├── node127.yml
└── zurich
├── zrh.node1.yml
├── zrh.node2.yml
└── zrh.node3.yml
**Examples**
``<saltclass_path>/nodes/lausanne/qls.node1.yml``
.. code-block:: yaml
environment: base
classes:
{% for class in ['default'] %}
- {{ class }}
{% endfor %}
- subsidiaries.{{ __grains__['id'].split('.')[0] }}
``<saltclass_path>/classes/default/init.yml``
.. code-block:: yaml
classes:
- default.users
- default.motd
states:
- openssh
pillars:
default:
network:
dns:
srv1: 192.168.0.1
srv2: 192.168.0.2
domain: example.com
ntp:
srv1: 192.168.10.10
srv2: 192.168.10.20
``<saltclass_path>/classes/subsidiaries/gnv.yml``
.. code-block:: yaml
pillars:
default:
network:
sub: Geneva
dns:
srv1: 10.20.0.1
srv2: 10.20.0.2
srv3: 192.168.1.1
domain: gnv.example.com
users:
adm1:
uid: 1210
gid: 1210
gecos: 'Super user admin1'
homedir: /srv/app/adm1
adm3:
uid: 1203
gid: 1203
gecos: 'Super user adm
Variable expansions:
Escaped variables are rendered as is - ``${test}``
Missing variables are rendered as is - ``${net:dns:srv2}``
.. code-block:: yaml
pillars:
app:
config:
dns:
srv1: ${default:network:dns:srv1}
srv2: ${net:dns:srv2}
uri: https://application.domain/call?\${test}
prod_parameters:
- p1
- p2
- p3
pkg:
- app-core
- app-backend
List override:
Not using ``^`` as the first entry will simply merge the lists
.. code-block:: yaml
pillars:
app:
pkg:
- ^
- app-frontend
**Known limitation**
Currently you can't have both a variable and an escaped variable in the same string as the escaped one will not be correctly rendered - '\${xx}' will stay as is instead of being rendered as '${xx}'
Newer PyWinRM Versions
----------------------
Versions of ``pywinrm>=0.2.1`` are finally able to disable validation of self
signed certificates. :ref:`Here<new-pywinrm>` for more information.
DigitalOcean
------------
The DigitalOcean driver has been renamed to conform to the company name. The
new driver name is ``digitalocean``. The old name ``digital_ocean`` and a
short one ``do`` will still be supported through virtual aliases, this is mostly
cosmetic.
Solaris Logical Domains In Virtual Grain
----------------------------------------
Support has been added to the ``virtual`` grain for detecting Solaris LDOMs
running on T-Series SPARC hardware. The ``virtual_subtype`` grain is
populated as a list of domain roles.
Lists of comments in state returns
----------------------------------
State functions can now return a list of strings for the ``comment`` field,
as opposed to only a single string.
This is meant to ease writing states with multiple or multi-part comments.
Beacon configuration changes
----------------------------
In order to remain consistent and to align with other Salt components such as states,
support for configuring beacons using dictionary based configuration has been deprecated
in favor of list based configuration. All beacons have a validation function which will
check the configuration for the correct format and only load if the validation passes.
- ``avahi_announce`` beacon
Old behavior:
.. code-block:: yaml
beacons:
avahi_announce:
run_once: True
servicetype: _demo._tcp
port: 1234
txt:
ProdName: grains.productname
SerialNo: grains.serialnumber
Comments: 'this is a test'
New behavior:
.. code-block:: yaml
beacons:
avahi_announce:
- run_once: True
- servicetype: _demo._tcp
- port: 1234
- txt:
ProdName: grains.productname
SerialNo: grains.serialnumber
Comments: 'this is a test'
- ``bonjour_announce`` beacon
Old behavior:
.. code-block:: yaml
beacons:
bonjour_announce:
run_once: True
servicetype: _demo._tcp
port: 1234
txt:
ProdName: grains.productname
SerialNo: grains.serialnumber
Comments: 'this is a test'
New behavior:
.. code-block:: yaml
beacons:
bonjour_announce:
- run_once: True
- servicetype: _demo._tcp
- port: 1234
- txt:
ProdName: grains.productname
SerialNo: grains.serialnumber
Comments: 'this is a test'
- ``btmp`` beacon
Old behavior:
.. code-block:: yaml
beacons:
btmp: {}
New behavior:
.. code-block:: yaml
beacons:
btmp: []
- ``glxinfo`` beacon
Old behavior:
.. code-block:: yaml
beacons:
glxinfo:
user: frank
screen_event: True
New behavior:
.. code-block:: yaml
beacons:
glxinfo:
- user: frank
- screen_event: True
- ``haproxy`` beacon
Old behavior:
.. code-block:: yaml
beacons:
haproxy:
- www-backend:
threshold: 45
servers:
- web1
- web2
- interval: 120
New behavior:
.. code-block:: yaml
beacons:
haproxy:
- backends:
www-backend:
threshold: 45
servers:
- web1
- web2
- interval: 120
- ``inotify`` beacon
Old behavior:
.. code-block:: yaml
beacons:
inotify:
/path/to/file/or/dir:
mask:
- open
- create
- close_write
recurse: True
auto_add: True
exclude:
- /path/to/file/or/dir/exclude1
- /path/to/file/or/dir/exclude2
- /path/to/file/or/dir/regex[a-m]*$:
regex: True
coalesce: True
New behavior:
.. code-block:: yaml
beacons:
inotify:
- files:
/path/to/file/or/dir:
mask:
- open
- create
- close_write
recurse: True
auto_add: True
exclude:
- /path/to/file/or/dir/exclude1
- /path/to/file/or/dir/exclude2
- /path/to/file/or/dir/regex[a-m]*$:
regex: True
- coalesce: True
- ``journald`` beacon
Old behavior:
.. code-block:: yaml
beacons:
journald:
sshd:
SYSLOG_IDENTIFIER: sshd
PRIORITY: 6
New behavior:
.. code-block:: yaml
beacons:
journald:
- services:
sshd:
SYSLOG_IDENTIFIER: sshd
PRIORITY: 6
- ``load`` beacon
Old behavior:
.. code-block:: yaml
beacons:
load:
1m:
- 0.0
- 2.0
5m:
- 0.0
- 1.5
15m:
- 0.1
- 1.0
emitatstartup: True
onchangeonly: False
New behavior:
.. code-block:: yaml
beacons:
load:
- averages:
1m:
- 0.0
- 2.0
5m:
- 0.0
- 1.5
15m:
- 0.1
- 1.0
- emitatstartup: True
- onchangeonly: False
- ``log`` beacon
Old behavior:
.. code-block:: yaml
beacons:
log:
file: <path>
<tag>:
regex: <pattern>
New behavior:
.. code-block:: yaml
beacons:
log:
- file: <path>
- tags:
<tag>:
regex: <pattern>
- ``network_info`` beacon
Old behavior:
.. code-block:: yaml
beacons:
network_info:
- eth0:
type: equal
bytes_sent: 100000
bytes_recv: 100000
packets_sent: 100000
packets_recv: 100000
errin: 100
errout: 100
dropin: 100
dropout: 100
New behavior:
.. code-block:: yaml
beacons:
network_info:
- interfaces:
eth0:
type: equal
bytes_sent: 100000
bytes_recv: 100000
packets_sent: 100000
packets_recv: 100000
errin: 100
errout: 100
dropin: 100
dropout: 100
- ``network_settings`` beacon
Old behavior:
.. code-block:: yaml
beacons:
network_settings:
eth0:
ipaddr:
promiscuity:
onvalue: 1
eth1:
linkmode:
New behavior:
.. code-block:: yaml
beacons:
network_settings:
- interfaces:
- eth0:
ipaddr:
promiscuity:
onvalue: 1
- eth1:
linkmode:
- ``proxy_example`` beacon
Old behavior:
.. code-block:: yaml
beacons:
proxy_example:
endpoint: beacon
```
New behavior:
```
beacons:
proxy_example:
- endpoint: beacon
- ``ps`` beacon
Old behavior:
.. code-block:: yaml
beacons:
ps:
- salt-master: running
- mysql: stopped
New behavior:
.. code-block:: yaml
beacons:
ps:
- processes:
salt-master: running
mysql: stopped
- ``salt_proxy`` beacon
Old behavior:
.. code-block:: yaml
beacons:
salt_proxy:
- p8000: {}
- p8001: {}
New behavior:
.. code-block:: yaml
beacons:
salt_proxy:
- proxies:
p8000: {}
p8001: {}
- ``sensehat`` beacon
Old behavior:
.. code-block:: yaml
beacons:
sensehat:
humidity: 70%
temperature: [20, 40]
temperature_from_pressure: 40
pressure: 1500
New behavior:
.. code-block:: yaml
beacons:
sensehat:
- sensors:
humidity: 70%
temperature: [20, 40]
temperature_from_pressure: 40
pressure: 1500
- ``service`` beacon
Old behavior:
.. code-block:: yaml
beacons:
service:
salt-master:
mysql:
New behavior:
.. code-block:: yaml
beacons:
service:
- services:
nginx:
onchangeonly: True
delay: 30
uncleanshutdown: /run/nginx.pid
- ``sh`` beacon
Old behavior:
.. code-block:: yaml
beacons:
sh: {}
New behavior:
.. code-block:: yaml
beacons:
sh: []
- ``status`` beacon
Old behavior:
.. code-block:: yaml
beacons:
status: {}
New behavior:
.. code-block:: yaml
beacons:
status: []
- ``telegram_bot_msg`` beacon
Old behavior:
.. code-block:: yaml
beacons:
telegram_bot_msg:
token: "<bot access token>"
accept_from:
- "<valid username>"
interval: 10
New behavior:
.. code-block:: yaml
beacons:
telegram_bot_msg:
- token: "<bot access token>"
- accept_from:
- "<valid username>"
- interval: 10
- ``twilio_txt_msg`` beacon
Old behavior:
.. code-block:: yaml
beacons:
twilio_txt_msg:
account_sid: "<account sid>"
auth_token: "<auth token>"
twilio_number: "+15555555555"
interval: 10
New behavior:
.. code-block:: yaml
beacons:
twilio_txt_msg:
- account_sid: "<account sid>"
- auth_token: "<auth token>"
- twilio_number: "+15555555555"
- interval: 10
- ``wtmp`` beacon
Old behavior:
.. code-block:: yaml
beacons:
wtmp: {}
New behavior:
.. code-block:: yaml
beacons:
wtmp: []
New requisites available in state compiler
------------------------------------------
- ``require_any``
The use of ``require_any`` demands that one of the required states executes before the
dependent state. The state containing the ``require_any`` requisite is defined as the
dependent state. The states specified in the ``require_any`` statement are defined as the
required states. If at least one of the required state's execution succeeds, the dependent state
will then execute. If at least one of the required state's execution fails, the dependent state
will not execute.
- ``watch_any``
The state containing the ``watch_any`` requisite is defined as the watching
state. The states specified in the ``watch_any`` statement are defined as the watched
states. When the watched states execute, they will return a dictionary containing
a key named "changes".
If the "result" of any of the watched states is ``True``, the watching state *will
execute normally*, and if all of them are ``False``, the watching state will never run.
This part of ``watch`` mirrors the functionality of the ``require`` requisite.
If the "result" of any of the watched states is ``True`` *and* the "changes"
key contains a populated dictionary (changes occurred in the watched state),
then the ``watch`` requisite can add additional behavior. This additional
behavior is defined by the ``mod_watch`` function within the watching state
module. If the ``mod_watch`` function exists in the watching state module, it
will be called *in addition to* the normal watching state. The return data
from the ``mod_watch`` function is what will be returned to the master in this
case; the return data from the main watching function is discarded.
If the "changes" key contains an empty dictionary, the ``watch`` requisite acts
exactly like the ``require`` requisite (the watching state will execute if
"result" is ``True``, and fail if "result" is ``False`` in the watched state).
- ``onchanges_any``
The ``onchanges_any`` requisite makes a state only apply one of the required states
generates changes, and if one of the watched state's "result" is ``True``. This can be
a useful way to execute a post hook after changing aspects of a system.
- ``onfail_any``
The ``onfail_any`` requisite allows for reactions to happen strictly as a response
to the failure of at least one other state. This can be used in a number of ways, such as
executing a second attempt to set up a service or begin to execute a separate
thread of states because of a failure.
The ``onfail_any`` requisite is applied in the same way as ``require_any`` and ``watch_any``:
Basic Slots support in states compiler
--------------------------------------
Slots extend the state syntax and allows you to do things right before the
state function is executed. So you can make a decision in the last moment right
before a state is executed.
Slot syntax looks close to the simple python function call. Here is a simple example:
.. code-block:: yaml
copy-some-file:
file.copy:
- name: __slot__:salt:test.echo(text=/tmp/some_file)
- source: __slot__:salt:test.echo(/etc/hosts)
Read more :ref:`here <slots-subsystem>`.
Deprecations
------------
Configuration Option Deprecations
=================================
- The ``requests_lib`` configuration option has been removed. Please use
``backend`` instead.
Profitbricks Cloud Updated Dependency
=====================================
The minimum version of the ``profitbrick`` python package for the ``profitbricks``
cloud driver has changed from 3.0.0 to 3.1.0.
Azure Cloud Updated Dependency
------------------------------
The azure sdk used for the ``azurearm`` cloud driver now depends on ``azure-cli>=2.0.12``
Module Deprecations
===================
The ``blockdev`` execution module has been removed. Its functions were merged
with the ``disk`` module. Please use the ``disk`` execution module instead.
The ``lxc`` execution module had the following changes:
- The ``dnsservers`` option to the ``cloud_init_interface`` function no longer
defaults to ``4.4.4.4`` and ``8.8.8.8``.
- The ``dns_via_dhcp`` option to the ``cloud_init_interface`` function defaults
to ``True`` now instead of ``False``.
The ``win_psget`` module had the following changes:
- The ``psversion`` function was removed. Please use ``cmd.shell_info`` instead.
The ``win_service`` module had the following changes:
- The ``config`` function was removed. Please use the ``modify`` function
instead.
- The ``binpath`` option was removed from the ``create`` function. Please use
``bin_path`` instead.
- The ``depend`` option was removed from the ``create`` function. Please use
``dependencies`` instead.
- The ``DisplayName`` option was removed from the ``create`` function. Please
use ``display_name`` instead.
- The ``error`` option was removed from the ``create`` function. Please use
``error_control`` instead.
- The ``group`` option was removed from the ``create`` function. Please use
``load_order_group`` instead.
- The ``obj`` option was removed from the ``create`` function. Please use
``account_name`` instead.
- The ``password`` option was removed from the ``create`` function. Please use
``account_password`` instead.
- The ``start`` option was removed from the ``create`` function. Please use
``start_type`` instead.
- The ``type`` option was removed from the ``create`` function. Please use
``service_type`` instead.
Runner Deprecations
===================
The ``manage`` runner had the following changes:
- The ``root_user`` kwarg was removed from the ``bootstrap`` function. Please
use ``salt-ssh`` roster entries for the host instead.
State Deprecations
==================
The ``archive`` state had the following changes:
- The ``tar_options`` and the ``zip_options`` options were removed from the
``extracted`` function. Please use ``options`` instead.
The ``cmd`` state had the following changes:
- The ``user`` and ``group`` options were removed from the ``run`` function.
Please use ``runas`` instead.
- The ``user`` and ``group`` options were removed from the ``script`` function.
Please use ``runas`` instead.
- The ``user`` and ``group`` options were removed from the ``wait`` function.
Please use ``runas`` instead.
- The ``user`` and ``group`` options were removed from the ``wait_script``
function. Please use ``runas`` instead.
The ``file`` state had the following changes:
- The ``show_diff`` option was removed. Please use ``show_changes`` instead.
Grain Deprecations
==================
For ``smartos`` some grains have been deprecated. These grains will be removed in Neon.
- The ``hypervisor_uuid`` has been replaced with ``mdata:sdc:server_uuid`` grain.
- The ``datacenter`` has been replaced with ``mdata:sdc:datacenter_name`` grain.
Minion Blackout
---------------
During a blackout, minions will not execute any remote execution commands,
except for :mod:`saltutil.refresh_pillar <salt.modules.saltutil.refresh_pillar>`.
Previously, support was added so that blackouts are enabled using a special
pillar key, ``minion_blackout`` set to ``True`` and an optional pillar key
``minion_blackout_whitelist`` to specify additional functions that are permitted
during blackout. This release adds support for using this feature in the grains
as well, by using special grains keys ``minion_blackout`` and
``minion_blackout_whitelist``.
Pillar Deprecations
-------------------
The legacy configuration for ``git_pillar`` has been removed. Please use the new
configuration for ``git_pillar``, which is documented in the external pillar module
for :mod:`git_pillar <salt.pillar.git_pillar>`.
Utils Deprecations
==================
The ``salt.utils.cloud.py`` file had the following change:
- The ``fire_event`` function now requires a ``sock_dir`` argument. It was previously
optional.
Other Miscellaneous Deprecations
================================
The ``version.py`` file had the following changes:
- The ``rc_info`` function was removed. Please use ``pre_info`` instead.
Warnings for moving away from the ``env`` option were removed. ``saltenv`` should be
used instead. The removal of these warnings does not have a behavior change. Only
the warning text was removed.
Sentry Log Handler
------------------
Configuring sentry raven python client via ``project``, ``servers``, ``public_key
and ``secret_key`` is deprecated and won't work with sentry clients > 3.0.
Instead, the ``dsn`` config param must be used.
Reference in New Issue View Git Blame Copy Permalink