* Module DOCUMENTATION should match argspec
Large update of many modules so that DOCUMENTATION option name and
aliases match those defined in the argspec.
Issues identified by https://github.com/ansible/ansible/pull/34809
In addition to many typos and missing aliases, the following notable
changes were made:
* Create `module_docs_fragments/url.py` for `url_argument_spec`
* `dellos*_command` shouldn't have ever had `waitfor` (was incorrectly copied)
* `ce_aaa_server_host.py` `s/raduis_server_type/radius_server_type/g`
* `Junos_lldp` enable should be part of `state`.
* Add VnicProfileMapping to register VM
Add vnic profile mappings to be supported in vm registration
* Add VnicProfileMapping to register template
Add vnic profile mappings to be supported in template registration
* Add reassign bad macs to register VM
Add reassign bad macs to be supported in vm registration.
* Add additional mappings params for VM registration
As part of the effort to support DR with oVirt
the "Register" operation is being added with a new mapping parameter
that describes the configuration of the registration.
The idea of supporting DR site to site in oVirt is to have 2 active
setups using storage replication between the primary setup and the
secondary setup.
Both setups will have active DCs, clusters, and hosts, although those
will not be identical.
The user can define a mapping which will be used to recover its setup.
Each mapping can be used to map any VM's attribute stored in the OVF
with its correlated entity.
For example, there could be a primary setup with a VM configured on cluster A.
We also keep an active secondary setup which only have cluster B.
Cluster B is compatible for that VM and in case of a DR scenario theoretically
the storage domain can be imported to the secondary setup and the use can
register the VM to cluster B.
In that case, we can automate the recovery process by defining a cluster mapping,
so once the entity will be registered its OVF will indicate it belongs to
cluster A but the mapping which will be sent will indicate that cluster B should
be valid for every thing that is configured on cluster A.
The engine should do the switch, and register the VM to cluster B in the secondary site.
Cluster mapping is just one example.
The following list describes the different mappings which were
introduced:
LUN mapping
Role mapping
Permissions mapping
Affinity group mapping
Affinity label mapping
Each mapping will be used for its specific OVF's data once the register operation
will take place in the engine.
* Add additional mappings params for Template registration
As part of the effort to support DR with oVirt
the "Register" operation is being added with a new mapping parameter
that describes the configuration of the registration.
The idea of supporting DR site to site in oVirt is to have 2 active
setups using storage replication between the primary setup and the
secondary setup.
Both setups will have active DCs, clusters, and hosts, although those
will not be identical.
The user can define a mapping which will be used to recover its setup.
Each mapping can be used to map any Template's attribute stored in the OVF
with its correlated entity.
For example, there could be a primary setup with a Template configured on cluster A.
We also keep an active secondary setup which only have cluster B.
Cluster B is compatible for that Template and in case of a DR scenario theoretically
the storage domain can be imported to the secondary setup and the use can
register the Template to cluster B.
In that case, we can automate the recovery process by defining a cluster mapping,
so once the entity will be registered its OVF will indicate it belongs to
cluster A but the mapping which will be sent will indicate that cluster B should
be valid for every thing that is configured on cluster A.
The engine should do the switch, and register the Template to cluster B in the
secondary site.
Cluster mapping is just one example.
The following list describes the different mappings which were
introduced:
Role mapping
Permissions mapping
Each mapping will be used for its specific OVF's data once the register operation
will take place in the engine.
* Add support for update OVF store
Add support for task of update OVF store in a storage domain.
* port elb_classic_facts to boto3
update module to use AnsibleAWSModule
* Add RETURN docs for elb_classic_lb_facts
* Remove superfluous exception handling around connection
Fix exit_json call and RETURN docs
oVirt modules support environment variables to be passed as
authentication details for connection. But ovirt_auth doesn't support
it. This patch add support for it.
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
* updates to azure_rm_sqlserver_facts
This fix refactor vmware_host module to use PyVmomi.
Also, handle SSLVerifyFault exception to get hostsystem's certificate
thumbprint.
Fixes: #20819
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* fix broken import
* ansible.module_utils.ec2.HAS_BOTO is already used
* wait_timeout parameter: use 'int' type
* wait is always True there
* doc: use formatting function
This adds a new module called 'vmware_host_config_facts' which
gathers facts about advanced configuration informantion about ESXi host or all
ESXi hosts from given cluster.
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* First pass at Terraform module
* Support output variables
* Support idempotent plan checks and external plan files
* Add check mode support
* Support custom statefile location & remove color-coding from shell output
* Docs
* Let missing statefile hard-fail
* Update to support present/planned/absent states
* PEP8
* Add docs & fix python 2.6 incompatibility
* PEP8
* PEP8
* Fix misleading cleanup/detach documentation
Current documentation is not concrete regarding desired
state of `detach` to make container be properly removed
with `cleanup`
Make `detach:false` be mentioned explicitly
* Tweak formatting
* adding azure_rm_postgresqlserver
* removed semicolon placed by mistake
* updates to azure_rm_postgresqlserver
* updates to azure_rm_postgresqlserver
* updates to azure_rm_postgresqlserver
* updates to azure_rm_postgresqlserver
* updates to azure_rm_postgresqlserver
* updates to azure_rm_postgresqlserver
* updates to azure_rm_postgresqlserver
* updates to azure_rm_postgresqlserver
* Updated dependencies
* Allow protection of certain keys during camel_to_snake
Create an `ignore_list` parameter that preserves the case
of the contents of certain dictionaries. Most valuable
for `tags` but other uses might arise.
* Port ec2_vpc_route_table to boto3
Update tests to reflect fixes in boto3.
* Add RETURN documentation to ec2_vpc_route_table
* Update DOCUMENTATION to be valid yaml
* Add check mode tests
* Add test for unexpected egress rule in default VPC
When passing rules_egress to ec2_group, the default
egress rule shouldn't be created (if `purge_rules_egress`)
is set. Test this.
* Respect egress rule defintions for default VPC groups
When passing rules_egress and purge_rules_egress, the
default egress rule should not be created
Fixes#34429
* Change AWS credential passing to be YAML anchors
Vastly simplify the AWS tasks by reducing the credentials to a YAML
block
* wrap boto3 connection in try/except and handle exception, add traceback import
* params dont need to be mutually exclusive and support check mode
* add check to set nacl_ids to empty list instead of None for boto3 func
* standard exception handling using traceback
* update current RETURN documentation which is incorrect.
* update logic to check for and return ipv6 cidr instead of ipv4 cidr in a NACL
* ignore default and reserved rule numbers - greater than 32766
* was breaking on non-standard protocol numbers - AWS nacls now support full list -1 to 254
* update port range and icmp type/code handling
* add some more detail to returned nacl entries
* Update exception handling to account for BotoCoreError
Remove exception handling from the boto connection creation since it does nothing
* rules numbers from 32767 to 65535 are reserved for internal AWS use so we ignore here
* Add some integration tests for ec2_asg.
* Remove exception handling from ec2_asg boto3 connection since it is handled in boto3_conn().
* Update test failure assertions
* Use yaml anchor for credentials and remove unnecessary dependencies.
* Move AWS boto3 module credentials tests to separate target
* Remove filters from tests
* cs_network_offering: streamline docs a bit
* cs_network_offering: format traffic type as returned by API
* cs_network_offering: rename return to service_offering_id
* cs_network_offering: ensure absent tests don't require anyting else than name
* cs_network_offering: beautify the default_egress_policy param
* cs_network_offering: test enabling/disabling with existing syntax
* cs_network_offering: rename service_provider_list to service_providers
* cs_network_offering: allow update max_connections
* cs_network_offering: extend returns
* cs_network_offering: extend tests with max_connections
* cloudstack: rename cs_serviceoffer to cs_service_offering
* adjust references to old name
* cs_service_offering: fix limitcpuuse not handled
* minor docs fix
* update changelog
This fix adds new module for managing various roles in
local ESXi server. Also, adds integration tests for this change.
Fixes: #27899
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* Use `AnsibleAWSModule`
* Update exception handling to use `fail_json_aws` and check
for `BotoCoreError` exceptions associated with bad connection
parameters.
* Remove connection creation exception handling as it does nothing.
This fix adds support for Virtual Machine resource allocation
information. This information is related to limit and reserve CPU and Memory
of virtual machine.
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
The DigitalOcean API v2 supports creating a block storage volume from
a previously-saved snapshot using a snapshot_id string. This module now
likewise supports creating a block storage volume using a valid
snapshot_id string.
When creating a block storage volume from a snapshot_id using the
DigitalOcean API, the region and size_gigabytes parameters are ignored.
Therefore, these parameters are likewise ignored when using this module.
However, as of this commit, they are still required fields. It may be
necessary to find some way to eliminate these requirements if the
snapshot_id parameter is included.
The DigitalOcean API v2 allows for creating a block storage volume from
a previously-saved snapshot using the snapshot_id parameter. If this
parameter is used, the region and size_gigabytes parameters are
disregarded. In order to avoid confusion, when the snapshot_id parameter
is included when creating a block storage volume using this module, the
region and block_size fields are overridden and changed to null values.
This should make it clear that these values are NOT being used if a
snapshot_id is present.
This module will still fail as before if neither the region/block_size
or snapshot_id is not present when creating a block storage volume.
Documentation has been updated to reflect these changes.
Remove examples from main section and make sure it is all in the examples. Should helpw ith the HTML breakage (it is too wide for most monitors).
tyop fixes too.
* disk can be mounted to vm
* add version added
* fix lint
* Fix some mirrors
* fix lint
* remove trailing space
* fix as review comment
* fix yaml
* fix
* fix
* Minor update to doc on unmounting disk from VM
* WIP adds network subnetting functions
* adds functions to convert between netmask and masklen
* adds functions to verify netmask and masklen
* adds function to dtermine network and subnet from address / mask pair
* network_common: add a function to get the first 48 bits in a IPv6 address.
ec2_group: only use network bits of a CIDR.
* Add tests for CIDRs with host bits set.
* ec2_group: add warning if CIDR isn't the networking address.
* Fix pep8.
* Improve wording.
* fix import for network utils
* Update tests to use pytest instead of unittest
* add test for to_ipv6_network()
* Fix PEP8
This fix check for ESXi server instance before proceeding
with managing local user. Also, adds integration tests for
this change.
Fixes: #32465
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* Port ec2_vpc_net to boto3 and add support to expand existing VPCs
* Add s-hertel as an author for ec2_vpc_net
* Update ec2_vpc_net test for new error triggered by lack of credentials
Fix backwards compatibility
Document new return value
* Fix pep8 and return documentation
* Remove boto usage from boto3 modules
* Remove region checking
boto3_conn now takes care of region checking and handles NoRegionError
exceptions with a standard message
boto3_conn also takes care of other connection exceptions too.
* Document boto3 as a requirement for ec2_eni_facts
* Return id of ENI in addition to network_interface_id. To be compatible to ec2_eni.
* Added documentation for the return values of the ec2_eni_facts module
* Fix typo in docs for ec2_eni_facts
`os.environ['DO_API_TOKEN']` raised a `KeyError` preventing the check
for `os.environ['DO_API_KEY]` from being executed. Fix this by failing
only if the api token isn't set.
* vmware_host: add reconnect and add_or_reconnect states
Add "reconnect" and "add_or_reconnect" choices for "state".
* reconnect: reconnect an esxi to a vcenter (imply it is present).
* add_or_reconnect: do the same but add the esxi if absent.
Also:
* tag the cluster_name as required (because it is).
* tag esxi_username and esxi_password as not required because
they aren't when the esxi isn't added.
* vmware_host: add + prepare/document integration tests
vmware_host module
Add integration test for the add part of "add_or_reconnect" state.
Prepare and document integration tests for the reconnect part
of "add_or_reconnect" state and "reconnect" and "absent" states.
Currently we can't test those states as ReconnectHost_Task (for
"reconnect") and EnterMaintenanceMode_Task (for "absent") aren't
implemented yet in vcsim (from vmware/govmomi)
* Readded vital return parameters to the ec2_asg that have been spared for no obvious reason
* Fix typo in ec2_asg docs
* Fixing another typo in ec2_asg docs.
* Added new module: Azure RM SQL Servers
* fixed sanity issues
* enabled integration test
* disabling test as Microsoft.Sql namespace not available
* renamed servers -> server
* updates as requested in the review
* udpates to sql server
* renamed module
* replaced has_key by in
* flattened results
* final updates to sql server
* updates as requested in the review
* added more tests
* deleting unexisting instance up front
* more tests
* another round of test changes
* removed uneeded test
This fix adds following:
* Documentation update
* Add support for DHCP
* Refactor code to use PyVmomi class
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* Refactor cloudscale API code
Move code common to all cloudscale cloud modules into a common base
class.
This is needed as a prepartion of the cloudscale_floating_ip module.
* cloudscale_floating_ip module
New cloud module to manage floating IPs on the cloudscale.ch IaaS
service.
Fix a bunch of things mentioned in the review.
Delete commented code from module. Add fix for vcsim not returning
uncommitted.
Add integration test.
Add changes suggested
* Bulk pep8 fixes - hand crafted
Fix by hand the remaining issues that autopep8 couldn't
* Next batch of hand crafted pep8 fixes
* Ignore W503
https://github.com/PyCQA/pycodestyle/pull/499
* Revert more of W503
* Fail when attempting to modify unmodifiable target group parameters
As you can't modify Port, Protocol or VPC id for a target group, fail
when this happens rather than pretending to do it.
One could argue that the target group could be recreated rather than
failing, but this has massive knock on implications to other resources
that depend on the TG (all ASGs would need to be updated, the ELB
listener would need to be updated, etc)
* Use `.get()` instead of direct dictionary access
* Added modules to create, delete, and describe EC2 Placement Groups.
* Remove unnecessary print statement
* Update to use boto3.
* De-linting
* Remove facts from this PR
* Update to newer method of handling Boto3 connections and exceptions.
* Futzing around with imports and HAS_BOTO3
* Fix up exception imports.
* Remove redundant default.
* Handle DryRunOperation errors appropriately.
* Remove redundant BOTO3 check.
* Use shorter licence declaration.
* Remove redundant HAS_BOTO3 import.
* Add AWSRetry decorators to API calls.
* Add new 'strategy' parameter to allow for cluster and spread PGs.
This fix adds following:
* hosts as argument spec
* cluster_name as argument spec
* refactor to use Pyvmomi class
* API refactor and documentation
Fixes: #18980
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* add acr
* acr
* acr
* fix a minor error
* add tests and fixes as suggested
* use get_mgmt_svc_client to enable bi
* fixes as zim suggested
* code refinement
* get acr instance before update
* fix a bug
* remove Classic sku as it's going to deprecate
* pass sanity test
* fix sanity
* fixes based on comments
* fix broken namespace unregistered error
* fix a bug
* fix a bug
* fix
* add register namespace
* fixes tests
* remove pdb
* add ci azure test
* removed required: false
* renamed acr -> containerregistry
* fixed spelling mistake
* fixed idempotence
* fix for idempotence was wrong
* fixed a few issues
* different log when exception different than ResourceNotFound
* self.fail instead of log
* updated sample return value
* added login server in results
* fixed else
* fixed line too long
* cmp -> __ne__
* fixed trailing whitespace
* output.changed is false when instance updated
* fix for state change
* one more update
* update instance only when location or sku has changed
* fixed syntax error
* fixed acr test
* just removed unnecessary comment to trigger ci
* update return value shape
* flatten `state`
* change credential dictionary state
* add credentials to integration tests
* Added module to describe EC2 Placement Groups.
* Update to newer way of handling Boto3 connections and exceptions.
* Futzing around with imports and HAS_BOTO3
* Fix up exception imports.
* Fix license block
* Correct task indenting in doc
* Added list-filtering example to docs
This fix make error message easily understable which is provided
by vcenter or ESXi server.
Fixes: #24672
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
This fix adds functionality to configure vmware guest parameter called
'maxMkconnections'. Also, added integration test for this change.
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
This fix adds functionality to support memory_reservation and
memory_reservation_lock while managing Virtual Machine.
Fixes: #20407
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
This fix adds missing comma in different modules.
This removes implicit string concatenation in given list.
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* Fix azure_rm_acs due to Azure breaking API change
* resolve code review feedbacks
* resolve test failures
* better name
* ignore case for vm size
* reformat code
* resolve code review feedbacks
Digital Ocean changed their api to no longer accept an integer as a resource_id. Fix adds changed data type (from integer to string).
Signed-off-by: Anthony Bond <bond.anthony@outlook.com>
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
This fix makes name and folder as required together
parameters in vmware_guest_snapshot.
Update integration tests for vmware_guest_snapshot.
Fixes: #22644
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
Fix adds esxi_hostname as Ansible module argument for user
to define ESXi hostname to deploy VMware vSwitch.
Fixes: #24647
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
All the values currently documented as return values are returned inside a 'result' key.
So if you registered the output of the task as 'output', then you would need to do 'output.result.zone_id' instead of 'output.zone_id'.
This commit fixes that so that you can do 'output.zone_id'.
This fix adds support for hardware parameter 'numCoresPerSocket'
in vmware_guest module. Also, adds integration tests for this change.
Fixes: #20406
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* check if need update for internal port
* validate port ip
* os_router modified in local
* my_os_router.py tested upto port not found
* tested need update
* default port attached with subnet getting deleted
* update happened with subnet and port id but rerun update=true?
* update working with portid converted subnetid for match checking
* tested and worked
* extra debug commnets cleaned up
* os_router with port ip tested fine
* deleted test files used wq for my development
* interface type changed for backward compatibility
* check if need update for internal port
* validate port ip
* os_router modified in local
* my_os_router.py tested upto port not found
* tested need update
* default port attached with subnet getting deleted
* update happened with subnet and port id but rerun update=true?
* update working with portid converted subnetid for match checking
* tested and worked
* extra debug commnets cleaned up
* os_router with port ip tested fine
* deleted test files used wq for my development
* check if need update for internal port
* validate port ip
* os_router modified in local
* my_os_router.py tested upto port not found
* tested need update
* default port attached with subnet getting deleted
* update happened with subnet and port id but rerun update=true?
* update working with portid converted subnetid for match checking
* tested and worked
* extra debug commnets cleaned up
* os_router with port ip tested fine
* deleted test files used wq for my development
* interface type changed for backward compatibility
* interface type changed for backward compatibility
* restoring requirement.txt which was deleted accidentally
* isinstance instead of type and white space removal
* trailing spaces removal
* multiple space after keyword 379,441
* fail.json interface type and deug msg changes
* test for membership should be 'not in'
This fix adds debugging details for 'no folder found' error.
This will allow to debug the user's setup easily.
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* add availability set for vm
* add test
* add import
* fix
* add test assert and output
* fix lint and version added
* add null check
* fix mirror
* fix mirror
* Improve error message in cloudscale_server module
Fix punctuation and add the full contents of "info" to the output in
case of failed API calls. This is useful in case of connection timeouts
and other error conditions where there is no response body available.
* Increase timeouts in cloudscale_server module
Increase the timeouts to not fail in case the API calls take a bit
longer than usual. The default timeout of fetch_url is 10s which is
quite short. Increase it to 30s. The timeout for waiting for a server
change is increased as well as it calls the API in a loop. Therefore
this value should be larger than the API timeout.
* Send API parameters as JSON in cloudscale_server module
Use JSON to send the POST data to the API instead of an urlencoded
string. Urlencoding is not really a good match for some Python
datatypes.
This fixes an issue when submitting a list of SSH keys which did not get
translated properly.
* Fix typo in cloudscale_server documentation
* cloudscale_sever: Replace timeout const by api_timeout param
Replace the static TIMEOUT_API constant by a user configurable
api_timeout parameter. Also eliminate the TIMEOUT_WAIT constant by
2*api_timeout. This means that the timeout to wait for server changes is
always double the timeout for API calls.
* Use Debian 9 image for cloudscale_server tests
* AWS S3: fix method call for fakes3 S3 backend
Fixes#33083
Signed-off-by: Marc-Aurèle Brothier <m@brothier.org>
* Auto append port based on proto
Signed-off-by: Marc-Aurèle Brothier <m@brothier.org>
This fix adds ability to set MAC address maunally. Before adding
any MAC address, the value is validated. If value is not valid, then
MAC address is set to vCenter generated MAC address.
Also, added integration tests for this change.
Fixes: #21161
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* Added eagerzeroedthick disk option to disk spec
* Updated docstrings for vmware_guest disk
* VMware: Integration test for disk type
Signed-off-by: Steve Jacobs <sjacobs@brokencrew.com>
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
John R Barker
maorlipchuk
Simone Tiraboschi
Abhijeet Kasurde
Matt Martz
Sloane Hertel
Ondra Machacek
Madhura-CSI
Daniel Mellado
Zim Kalinowski
René Moser
dpassante
Pilou
reversecipher
charliejllewellyn
Ryan Brown
Yury Tsarev
Zhijun Zhao
Zhijun Zhao
Will Thames
Daniel Shepherd
Vinay Dandekar
Patrick Marques
ZhijunZhao
Takuya Sato
Mohit Kumar
Jacob F. Grant
Will Thames
Igor (Tsigankov) Eyrich
Ted Timmons
Gustavo Maia
Rob
Yuwei Zhou
radixdlt
Kim Egede Jakobsen
Gobin Sougrakpam
Brian Lamar
Øyvind Saltvik
Ethan Devenport
Prasad Katti
Matthew Staebler
Mamad Purbo
Stefan Horning
tchernomax
Katerina Koukiou
pmargara
Gurchet Rai
Will Weber
jglf
Tal Zur
Gaudenz Steinlin
Tim Rightnour
mikedlr
Justin Menga
Matt Doller
Reza Jelveh
Brad Macpherson
Deepakkothandan
Matt Davis
Denis Chapligin
Christopher Bradley
Kaz Cheng
ABond
Daniel Temme
Nicolas DOUILLET
Harris Dimitriou
BhujayKumarBhatta
Vincent Bernat
Brad Chamberlain
Simon Fullick
Nijin Ashok
Remon van de Kamp
Matt Clay
Marc-Aurèle Brothier
Jonas Boesch
mldmld68
Steve Jacobs
Joerg Kastning
Philippe Dellaert