{Redis} Add support for 2022-06-01 API version

[austintolani]

Related command

az redis

Description

This PR adds supports for a new API version for Azure Cache for Redis (2022-06-01). Specifically, these changes include:

• Upgrading azure-mgmt-redis package to new version (14.1.0), which updates all requests to use the new 2022-06-01 API version
• Add a new optional parameter, --preferred-data-archive-method to az redis import and az redis export operation.
• Linked server deletion (az redis server-link delete) is now a long running operation so it now calls begin_delete instead of delete. Linked server deletion behavior is unchanged for the user.
• Updating a cache (az redis update) is now a long running operation so it now calls begin_update instead of update. Implemented a customer setter function so as not to break current behavior.
• A linked server has two new properties: geoReplicatedPrimaryHostName and primaryHostName.
• Relevant updates to help text
• Relevant updates to tests

Testing Guide

Updating a cache (example below shows updating the SKU):

az redis update --name MyRedisCache --resource-group MyResourceGroup --set "sku.capacity"="2"

Linked server operations:

az redis create -n {primaryName} -g {rg} -l {location} --sku {sku} --vm-size {size} # create primary
az redis create -n {secondaryName} -g {rg} -l {location} --sku {sku} --vm-size {size} # create secondary

az redis server-link create -n {primaryName} -g {rg} --replication-role Secondary --server-to-link {secondaryName} # link servers

az redis server-link show -n {primaryName} -g {rg} --linked-server-name {secondaryName} # see link details

az redis server-link delete -n {primaryName} -g {rg} --linked-server-name {secondaryName} # remove link

History Notes

[Redis] az redis import/export: Add new optional parameter --preferred-data-archive-method
[Redis] az redis server-link: Linked server has two new properties: geoReplicatedPrimaryHostName and primaryHostName

---

This checklist is used to make sure that common guidelines for a pull request are followed.

• The PR title and description has followed the guideline in Submitting Pull Requests.

• I adhere to the Command Guidelines.

• I adhere to the Error Handling Guidelines.

[yonzhan]

Redis

[evelyn-ys]

There're still two tests failing:

FAILED src/azure-cli/azure/cli/command_modules/network/tests/latest/test_private_endpoint_commands.py::NetworkPrivateLinkAzureCacheforRedisScenarioTest::test_private_endpoint_connection_acfr
FAILED src/azure-cli/azure/cli/command_modules/network/tests/latest/test_private_endpoint_commands.py::NetworkPrivateLinkAzureCacheforRedisScenarioTest::test_private_link_resource_acfr

Also pls fix the credential scanner failure

[austintolani]

@evelyn-ys I have fixed the failing tests and credential scanning issue.

[jiasli]

According to the definition of azure.core.polling._poller.LROPoller.result:

https://github.com/Azure/azure-sdk-for-python/blob/d678ef6e5ebe71bcc7bb7f3665d76576437f0121/sdk/core/azure-core/azure/core/polling/_poller.py#L242-L252

    def result(self, timeout: Optional[float] = None) -> PollingReturnType_co:
        """Return the result of the long running operation, or
        the result available after the specified timeout.


        :param float timeout: Period of time to wait before getting back control.
        :returns: The deserialized resource of the long running operation, if one is available.
        :rtype: any or None
        :raises ~azure.core.exceptions.HttpResponseError: Server problem with the query.
        """
        self.wait(timeout)
        return self._polling_method.resource()

The 0 in result(0) makes the LROPoller return immediately, causing intermittent CI failure:

https://dev.azure.com/azclitools/public/_build/results?buildId=121462&view=logs&j=038813f6-b85b-5e37-acf4-30b28352865f&t=48a22e66-ceab-5c7f-d00b-cb2046709dfd

2024-01-12T08:59:22.0016933Z =================================== FAILURES ===================================
2024-01-12T08:59:22.0017289Z ___________________ RedisCacheTests.test_redis_cache_update ____________________
2024-01-12T08:59:22.0019920Z [gw2] linux -- Python 3.11.5 /opt/az/bin/python3
2024-01-12T08:59:22.0020425Z /opt/az/lib/python3.11/site-packages/azure/cli/testsdk/base.py:145: in tearDown
2024-01-12T08:59:22.0020723Z     super(ScenarioTest, self).tearDown()
2024-01-12T08:59:22.0020984Z _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
2024-01-12T08:59:22.0021110Z 
2024-01-12T08:59:22.0021390Z self = <command_modules.redis.tests.latest.test_redis_scenario.RedisCacheTests testMethod=test_redis_cache_update>
2024-01-12T08:59:22.0021580Z 
2024-01-12T08:59:22.0021788Z     def tearDown(self):
2024-01-12T08:59:22.0022010Z         os.environ = self.original_env
2024-01-12T08:59:22.0022239Z         # Autorest.Python 2.x
2024-01-12T08:59:22.0022527Z         assert not [t for t in threading.enumerate() if t.name.startswith("AzureOperationPoller")], \
2024-01-12T08:59:22.0022994Z             "You need to call 'result' or 'wait' on all AzureOperationPoller you have created"
2024-01-12T08:59:22.0023274Z         # Autorest.Python 3.x
2024-01-12T08:59:22.0027370Z >       assert not [t for t in threading.enumerate() if t.name.startswith("LROPoller")], \
2024-01-12T08:59:22.0028218Z             "You need to call 'result' or 'wait' on all LROPoller you have created"
2024-01-12T08:59:22.0028728Z E       AssertionError: You need to call 'result' or 'wait' on all LROPoller you have created
2024-01-12T08:59:22.0028890Z 
2024-01-12T08:59:22.0029292Z /opt/az/lib/python3.11/site-packages/azure/cli/testsdk/scenario_tests/base.py:160: AssertionError
2024-01-12T08:59:22.0029744Z ------------- generated xml file: /azure_cli_test_result/redis.xml -------------
2024-01-12T08:59:22.0030074Z =========================== short test summary info ============================
2024-01-12T08:59:22.0030422Z FAILED tests/latest/test_redis_scenario.py::RedisCacheTests::test_redis_cache_update

No timeout should be given. result() should be used instead, such as:

azure-cli/src/azure-cli/azure/cli/command_modules/redis/custom.py

Line 231 in 573410a

redis_resourse = client.begin_update(resource_group_name, cache_name, update_params).result()

[evelyn-ys]

Per the PR description:

Updating a cache (az redis update) is now a long running operation so it now calls begin_update instead of update. Implemented a customer setter function so as not to break current behavior.

I think it's be design to make the operation returning immediately. But they should use sdk_no_wait instead of result(0)

[jiasli]

sdk_no_wait turns off polling:

azure-cli/src/azure-cli-core/azure/cli/core/util.py

Lines 707 to 710 in a5198b5

	def sdk_no_wait(no_wait, func, *args, **kwargs):
	if no_wait:
	kwargs.update({'polling': False})
	return func(*args, **kwargs)

but result(0) doesn't. result(0) will make the result returned immediately, but SDK still keeps polling in the background.

[austintolani]

Hi @jiasli and @evelyn-ys, thanks for your work identifying this issue. Is the expectation that the service team make this change or is this going to be handled by the CLI team?

[evelyn-ys]

appreciated that @austintolani you can create another PR to fix!

[austintolani]

Okay thanks @evelyn-ys. I am no longer on the Redis team so I don't think it's appropriate for me to make this change but I have forwarded this information to the Redis sub-team that owns the CLI.