fix: Major refactoring of Polling, Retry and Timeout logic #462

vam-google · 2022-10-19T02:43:38Z

If you are a user of python-api-core and experience polling timing out after ~15 minutes (900s) in your python code after this change, please make sure that instead of calling PollingFuture.result() you call it with additonal timeout argument like this: PollingFuture.result(timeout = <desired timeout in seconds>) or PollingFuture.result(timeout = None) (for infinite timeout, but infinite timeouts are strongly discouraged) for your code to stop timing out at 900s.

The core libraries cannot and should not have infitine polling as as default behavior. The fact that it had it like that in python for years (unlike other GCP-supported languages) was a bug, as it contradicts original cross-language LRO methods design in GAPIC libraries and in general users are not supposed to be put into an infinite-loop scenario implicitly. If an LRO can actually ran for hours or even days and it is Ok, then users are expected to acknowledge this fact by providing the huge timeouts explicitly.

This is in response to https://freeman.vc/notes/aws-vs-gcp-reliability-is-wildly-different, which triggered an investigation of the whole Polling/Retry/Timeout behavior in Python GAPIC clients and revealed many fundamental flaws in its implementation.

To properly describe the refactoring in this PR we need to stick to a rigorous terminology, as vague definitions of retries, timeouts, polling and related concepts seems to be the main source of the present bugs and overall confusion among both groups: users of the library and creators of the library. Please check the updated (in this PR) documentation of the google.api_core.retry.Retry class and the google.api_core.future.polling.Polling.result() method for the proper definitions and context.

Note, the overall semantics around Polling, Retry and Timeout remains quite confusing even after refactoring (although it is now more or less rigorously defined), but it was as clean as I could make it while still maintaining backward compatibility of the whole library.

The quick summary of the changes in this PR:

Properly define and fix the application of Deadline and Timeout concepts. Please check the updated documentation for the google.api_core.retry.Retry class for the actual definitions. Originally the deadline has been used to represent timeouts conflating the two concepts. As result this PR replaces deadline arguments with timeout ones in as backward-compatible manner as possible (i.e. backward compatible in all practical applications).
Properly define RPC Timeout, Retry Timeout and Polling Timeout and how a generic Timeout concept (aka Logical Timeout) is mapped to one of those depending on the context. Please check google.api_core.retry.Retry class documentation for details.
Properly define and fix the application of Retry and Polling concepts. Please check the updated documentation for google.api_core.future.polling.PollingFuture.result() for details.
Separate retry and polling configurations for Polling future, as these are two different concepts (although both operating on Retry class). Originally both retry and polling configurations were controlled by a single retry parameter, merging configuration regarding how "rpc error responses" and how "operation not completed" responses are supposed to be handled.
For the following config properties - Retry (including Retry Timeout), Polling (including Polling Timeout) and RPC Timeout - fix and properly define how each of the above properties gets configured and which config gets precedence in case of a conflict (check PollingFuture.result() method documentation for details). Each of those properties can be specified as follows: directly provided by the user for each call, specified during gapic generation time from config values in grpc_service_config.json file (for Retry and RPC Timeout) and gapic.yaml file (for Polling Timeout), or be provided as a hard-coded basic default values in python-api-core library itself. This alo includes fixing the per-call polling config propagation logic (the polling/retry configs supplied to PollingFuture.result() used to be ignored for actual call).
Deprecate ExponentialTimeout, ConstantTimeout and related logic as those are outdated concepts and are not consistent with the other GAPIC Languages. Replace it with TimeToDeadlineTimeout to be consistent with how the rest of the languages do it.
Deprecate google.api_core.operations_v1.config as it is an outdated concept and self-inconsistent (as all gapic clients provide configuraiton in code). The configs are directly provided in code instead.
Switch randomized delay calculation from delay being treated as expected value for randomized_delay to delay being treated as maximum value for randomized_delay (i.e. the new expected valud for randomized_delay is delay / 2). See the exponential_sleep_generator() method implementation for details. This is needed to make Python implementation of retries and polling exponential backoff consistent with the rest of GAPIC languages. Also fix the uncontrollable growth of delay value (since it is a subject of exponential growth, the delay value was quickly reaching "infinity" value, and the whole thing was not failing simply due to python being a very forgiving language which forgives multiplying "infinity" by a number (inf * number = inf) binstead of simply overflowing to a (most likely) negative number). Also essentially rollback the 52f12af change, since that is inconsistent with the other languages and damages uniform distibution of retry delays artificially shifting their concentration towards the end of timeout.
Fix url construction in OperationsRestTransport. Without this fix the polling logic for REST transport was completely broken (is not affecting Compute client, as that one has custom LRO).
Last but not least: change the default values for Polling logic to be the following: initial=1.0 (same as before), maximum=20.0 (was 60), multiplier=1.5 (was 2.0), timeout=900 (was 120, but due to timeout resolution logic was actually None (i.e. infinity)). This, in conjunction with changed calculation of randomized delay (i.e. its expected value now being delay / 2) overall makes polling logic much less aggressive in terms of increasing delays between each polling iteration, making LRO return much earlier for users on average, but still keeping a healthy balance between strain put on both client and server by polling and responsiveness of LROs for user.

*The design doc summarising all the changes and reasons for them is in progress.

In addition to the timeout/retry fixes, this PR has some other non-related technical fixes:

Fix and improve code coverage (and explicitly disables lines which are not supposed to be covered)
Define Python 3.10 as the default python version used in CI.
Upgrade to Sphinx 4.2.0, as it is the one supporting Python 3.10

This is in response to https://freeman.vc/notes/aws-vs-gcp-reliability-is-wildly-different, which triggered an investigation of the whole Polling/Retry/Timeout behavior in Python GAPIC clients and revealed many fundamental flaws in its implementaiton. To properly describe the refactoring this PR does we need to stick to a rigorous terminology, as vague definitions of retries, timeouts, polling and related concepts seems to be the main source of the present bugs and overal confusion among both groups: users of the library and creators of the library. Please check the documentation of the `google.api_core.retry.Retry` class the `google.api_core.future.polling.Polling.result()` method for the proper definitions and context. Note, the overall semantics around Polling, Retry and Timeout remains quite confusing even after refactoring (although it is now more or less rigorously defined), but it was clean as I could make it while still maintaining backward compatibility of the whole library. The quick summary of the changes in this PR: 1) Properly define and fix the application of Deadline and Timeout concepts. Please check the updated documentation for the `google.api_core.retry.Retry` class for the actual definitions. Originally the `deadline` has been used to represent timeouts conflating the two concepts. As result this PR replaces `deadline` arguments with `timeout` ones in as backward-compatible manner as possible (i.e. backward compatible in all practical applications). 2) Properly define RPC Timeout, Retry Timeout and Pollint Timeout and how a generic Timeout concept (aka Logical Timeout) is mapped to one of those depending on the context. Please check `google.api_core.retry.Retry` class documentation for details. 3) Properly define and fix the application of Retry and Polling concepts. Please check the updated documentation for `google.api_core.future.polling.PollingFuture.result()` for details. 4) Separate `retry` and `polling` configurations for Polling future, as these are two different concepts (although both operating on `Retry` class). Originally both retry and polling configurations were controlled by a single `retry` parameter, merging configuration regarding how "rpc error responses" and how "operation not completed" responses are supposed to be handled. 5) For the following config properties - `Retry (including `Retry Timeout`), `Polling` (including `Polling Timeout`) and `RPC Timeout` - fix and properly define how each of the above properties gets configured and which config gets precedence in case of a conflict (check `PollingFuture.result()` method documentation for details). Each of those properties can be specified as follows: directly provided by the user for each call, specified during gapic generation time from config values in `grpc_service_config.json` file (for Retry and RPC Timeout) and `gapic.yaml` file (for Polling), or be provided as a hard-coded basic default values in python-api-core library itself. 6) Fix the per-call polling config propagation logic (the polling/retry configs supplied to `PollingFuture.result()` used to be ignored for actual call). 7) Deprecate the usage of `deadline` terminology in the whole library and backward-compatibly replace it with timeout. This is essential as what has been called "deadline" in this library was actually "timeout" as it is defined in `google.api_core.retry.Retry` class documentation. 8) Deprecate `ExponentialTimeout`, `ConstantTimeout` and related logic as those are outdated concepts and are not consistent with the other GAPIC Languages. Replace it with `TimeToDeadlineTimeout` to be consistent with how the rest of the languages do it. 9) Deprecate `google.api_core.operations_v1.config` as it is an outdated concept and self-inconsistent (as all gapic clients provide configuraiton in code). The configs are directly provided in code instead. 10) Switch randomized delay calculation from `delay` being treated as expected value for randomized_delay to `delay` being treated as maximum value for `randomized_delay` (i.e. the new expected valud for `randomized_delay` is `delay / 2`). See the `exponential_sleep_generator()` method implementation for details. This is needed to make Python implementation of retries and polling exponential backoff consistent with the rest of GAPIC languages. Also fix the uncontrollable growth of `delay` value (since it is a subject of exponential growth, the `delay` value was quickly reaching "infinity" value, and the whole thing was not failing simply due to python being a very forgiving language which forgives multiplying "infinity" by a number (`inf * number = inf`) binstead of simply overflowing to a (most likely) negative number). 11) Fix url construction in `OperationsRestTransport`. Without this fix the polling logic for REST transport was completely broken (is not affecting Compute client, as that one has custom LRO). 12) Las but not least: change the default values for Polling logic to be the following: `initial=1.0` (same as before), `maximum=20.0` (was `60`), `multiplier=1.5` (was `2.0`), `timeout=900` (was `120`, but due to timeout resolution logic was actually None (i.e. infinity)). This, in conjunction with changed calculation of randomized delay (i.e. its expected value now being `delay / 2`) overall makes polling logic much less aggressive in terms of increasing delays between each polling iteration, making LRO return much earlier for users on average, but still keeping a healthy balance between strain put on both client and server by polling and responsiveness of LROs for user. *The design doc summarising all the changes and reasons for them is in progress.

Also fix LRO for REST transport. This PR makes generated gapics appeciate timeout values from grpc_service_config.json instead of overriding them with None (which means no timeout) It is basically a direct fix for googleapis#1477. This PR depends on googleapis/python-api-core#462, and expects `setup.py.j2` templates to be updated after googleapis/python-api-core#462 gets pushed and released with new version.

google/api_core/future/polling.py

google/api_core/extended_operation.py

google/api_core/future/polling.py

vam-google · 2022-11-01T17:05:34Z

@atulep Addressed your comments PTAL

tests/asyncio/gapic/test_method_async.py

tests/asyncio/test_retry_async.py

tests/unit/test_retry.py

tests/unit/gapic/test_method.py

alexander-fenster · 2022-11-03T18:07:01Z

@vam-google notification test

vam-google · 2022-11-04T17:52:10Z

@vam-google test

vchudnov-g

I have not looked carefully at the tests yet, but so far this looks great. I will be running some manual tests to complement what Tony and Aza have been doing.

Most of the comments are documentation tweaks, so hopefully you can just accept them on GitHub and pull them into your local repo before making any other changes.

vchudnov-g · 2022-11-07T22:03:55Z

google/api_core/future/polling.py

    _OperationNotComplete,
    exceptions.TooManyRequests,
    exceptions.InternalServerError,
    exceptions.BadGateway,
    exceptions.ServiceUnavailable,
 )
-DEFAULT_RETRY = retry.Retry(predicate=RETRY_PREDICATE)
+
+# DEPRECATED, use DEFAULT_POLLING to configure LRO polling logic. Construct


Could you be explicit about how to use DEFAULT_RETRY? It's not immediately clear how that relates to this sentence.
Also, maybe being more explicit about "baseline".

Maybe something like "Construct the Retry object using DEFAULT_RETRY as a baseline and then modifying specific parameters as needed"

The baseliine part was there before. Just keeping it as is, as the current recommendation is simply to not used that at all for anything.

As for being explicit on usage of DEFAULT_RETRY, it is deprecated, so it should not be used.

google/api_core/future/polling.py

google/api_core/retry_async.py

google/api_core/timeout.py

tests/unit/test_general_helpers.py

Co-authored-by: Victor Chudnovsky <vchudnov@google.com>

* fix: Major refactoring and fix for Polling, Retry and Timeout logic This is in response to https://freeman.vc/notes/aws-vs-gcp-reliability-is-wildly-different, which triggered an investigation of the whole Polling/Retry/Timeout behavior in Python GAPIC clients and revealed many fundamental flaws in its implementaiton. To properly describe the refactoring this PR does we need to stick to a rigorous terminology, as vague definitions of retries, timeouts, polling and related concepts seems to be the main source of the present bugs and overal confusion among both groups: users of the library and creators of the library. Please check the documentation of the `google.api_core.retry.Retry` class the `google.api_core.future.polling.Polling.result()` method for the proper definitions and context. Note, the overall semantics around Polling, Retry and Timeout remains quite confusing even after refactoring (although it is now more or less rigorously defined), but it was clean as I could make it while still maintaining backward compatibility of the whole library. The quick summary of the changes in this PR: 1) Properly define and fix the application of Deadline and Timeout concepts. Please check the updated documentation for the `google.api_core.retry.Retry` class for the actual definitions. Originally the `deadline` has been used to represent timeouts conflating the two concepts. As result this PR replaces `deadline` arguments with `timeout` ones in as backward-compatible manner as possible (i.e. backward compatible in all practical applications). 2) Properly define RPC Timeout, Retry Timeout and Pollint Timeout and how a generic Timeout concept (aka Logical Timeout) is mapped to one of those depending on the context. Please check `google.api_core.retry.Retry` class documentation for details. 3) Properly define and fix the application of Retry and Polling concepts. Please check the updated documentation for `google.api_core.future.polling.PollingFuture.result()` for details. 4) Separate `retry` and `polling` configurations for Polling future, as these are two different concepts (although both operating on `Retry` class). Originally both retry and polling configurations were controlled by a single `retry` parameter, merging configuration regarding how "rpc error responses" and how "operation not completed" responses are supposed to be handled. 5) For the following config properties - `Retry (including `Retry Timeout`), `Polling` (including `Polling Timeout`) and `RPC Timeout` - fix and properly define how each of the above properties gets configured and which config gets precedence in case of a conflict (check `PollingFuture.result()` method documentation for details). Each of those properties can be specified as follows: directly provided by the user for each call, specified during gapic generation time from config values in `grpc_service_config.json` file (for Retry and RPC Timeout) and `gapic.yaml` file (for Polling), or be provided as a hard-coded basic default values in python-api-core library itself. 6) Fix the per-call polling config propagation logic (the polling/retry configs supplied to `PollingFuture.result()` used to be ignored for actual call). 7) Deprecate the usage of `deadline` terminology in the whole library and backward-compatibly replace it with timeout. This is essential as what has been called "deadline" in this library was actually "timeout" as it is defined in `google.api_core.retry.Retry` class documentation. 8) Deprecate `ExponentialTimeout`, `ConstantTimeout` and related logic as those are outdated concepts and are not consistent with the other GAPIC Languages. Replace it with `TimeToDeadlineTimeout` to be consistent with how the rest of the languages do it. 9) Deprecate `google.api_core.operations_v1.config` as it is an outdated concept and self-inconsistent (as all gapic clients provide configuraiton in code). The configs are directly provided in code instead. 10) Switch randomized delay calculation from `delay` being treated as expected value for randomized_delay to `delay` being treated as maximum value for `randomized_delay` (i.e. the new expected valud for `randomized_delay` is `delay / 2`). See the `exponential_sleep_generator()` method implementation for details. This is needed to make Python implementation of retries and polling exponential backoff consistent with the rest of GAPIC languages. Also fix the uncontrollable growth of `delay` value (since it is a subject of exponential growth, the `delay` value was quickly reaching "infinity" value, and the whole thing was not failing simply due to python being a very forgiving language which forgives multiplying "infinity" by a number (`inf * number = inf`) binstead of simply overflowing to a (most likely) negative number). 11) Fix url construction in `OperationsRestTransport`. Without this fix the polling logic for REST transport was completely broken (is not affecting Compute client, as that one has custom LRO). 12) Las but not least: change the default values for Polling logic to be the following: `initial=1.0` (same as before), `maximum=20.0` (was `60`), `multiplier=1.5` (was `2.0`), `timeout=900` (was `120`, but due to timeout resolution logic was actually None (i.e. infinity)). This, in conjunction with changed calculation of randomized delay (i.e. its expected value now being `delay / 2`) overall makes polling logic much less aggressive in terms of increasing delays between each polling iteration, making LRO return much earlier for users on average, but still keeping a healthy balance between strain put on both client and server by polling and responsiveness of LROs for user. *The design doc summarising all the changes and reasons for them is in progress. * fix ci failures (mainly sphinx errors) * remove unused code * fix typo * Pin pytest version to <7.2.0 * reformat code * address pr feedback * address PR feedback * address pr feedback * Update google/api_core/future/polling.py Co-authored-by: Victor Chudnovsky <vchudnov@google.com> * Apply documentation suggestions from code review Co-authored-by: Victor Chudnovsky <vchudnov@google.com> * Address PR feedback Co-authored-by: Victor Chudnovsky <vchudnov@google.com>

…ch (#474) * fix: Major refactoring of Polling, Retry and Timeout logic (#462) * fix: Major refactoring and fix for Polling, Retry and Timeout logic This is in response to https://freeman.vc/notes/aws-vs-gcp-reliability-is-wildly-different, which triggered an investigation of the whole Polling/Retry/Timeout behavior in Python GAPIC clients and revealed many fundamental flaws in its implementaiton. To properly describe the refactoring this PR does we need to stick to a rigorous terminology, as vague definitions of retries, timeouts, polling and related concepts seems to be the main source of the present bugs and overal confusion among both groups: users of the library and creators of the library. Please check the documentation of the `google.api_core.retry.Retry` class the `google.api_core.future.polling.Polling.result()` method for the proper definitions and context. Note, the overall semantics around Polling, Retry and Timeout remains quite confusing even after refactoring (although it is now more or less rigorously defined), but it was clean as I could make it while still maintaining backward compatibility of the whole library. The quick summary of the changes in this PR: 1) Properly define and fix the application of Deadline and Timeout concepts. Please check the updated documentation for the `google.api_core.retry.Retry` class for the actual definitions. Originally the `deadline` has been used to represent timeouts conflating the two concepts. As result this PR replaces `deadline` arguments with `timeout` ones in as backward-compatible manner as possible (i.e. backward compatible in all practical applications). 2) Properly define RPC Timeout, Retry Timeout and Pollint Timeout and how a generic Timeout concept (aka Logical Timeout) is mapped to one of those depending on the context. Please check `google.api_core.retry.Retry` class documentation for details. 3) Properly define and fix the application of Retry and Polling concepts. Please check the updated documentation for `google.api_core.future.polling.PollingFuture.result()` for details. 4) Separate `retry` and `polling` configurations for Polling future, as these are two different concepts (although both operating on `Retry` class). Originally both retry and polling configurations were controlled by a single `retry` parameter, merging configuration regarding how "rpc error responses" and how "operation not completed" responses are supposed to be handled. 5) For the following config properties - `Retry (including `Retry Timeout`), `Polling` (including `Polling Timeout`) and `RPC Timeout` - fix and properly define how each of the above properties gets configured and which config gets precedence in case of a conflict (check `PollingFuture.result()` method documentation for details). Each of those properties can be specified as follows: directly provided by the user for each call, specified during gapic generation time from config values in `grpc_service_config.json` file (for Retry and RPC Timeout) and `gapic.yaml` file (for Polling), or be provided as a hard-coded basic default values in python-api-core library itself. 6) Fix the per-call polling config propagation logic (the polling/retry configs supplied to `PollingFuture.result()` used to be ignored for actual call). 7) Deprecate the usage of `deadline` terminology in the whole library and backward-compatibly replace it with timeout. This is essential as what has been called "deadline" in this library was actually "timeout" as it is defined in `google.api_core.retry.Retry` class documentation. 8) Deprecate `ExponentialTimeout`, `ConstantTimeout` and related logic as those are outdated concepts and are not consistent with the other GAPIC Languages. Replace it with `TimeToDeadlineTimeout` to be consistent with how the rest of the languages do it. 9) Deprecate `google.api_core.operations_v1.config` as it is an outdated concept and self-inconsistent (as all gapic clients provide configuraiton in code). The configs are directly provided in code instead. 10) Switch randomized delay calculation from `delay` being treated as expected value for randomized_delay to `delay` being treated as maximum value for `randomized_delay` (i.e. the new expected valud for `randomized_delay` is `delay / 2`). See the `exponential_sleep_generator()` method implementation for details. This is needed to make Python implementation of retries and polling exponential backoff consistent with the rest of GAPIC languages. Also fix the uncontrollable growth of `delay` value (since it is a subject of exponential growth, the `delay` value was quickly reaching "infinity" value, and the whole thing was not failing simply due to python being a very forgiving language which forgives multiplying "infinity" by a number (`inf * number = inf`) binstead of simply overflowing to a (most likely) negative number). 11) Fix url construction in `OperationsRestTransport`. Without this fix the polling logic for REST transport was completely broken (is not affecting Compute client, as that one has custom LRO). 12) Las but not least: change the default values for Polling logic to be the following: `initial=1.0` (same as before), `maximum=20.0` (was `60`), `multiplier=1.5` (was `2.0`), `timeout=900` (was `120`, but due to timeout resolution logic was actually None (i.e. infinity)). This, in conjunction with changed calculation of randomized delay (i.e. its expected value now being `delay / 2`) overall makes polling logic much less aggressive in terms of increasing delays between each polling iteration, making LRO return much earlier for users on average, but still keeping a healthy balance between strain put on both client and server by polling and responsiveness of LROs for user. *The design doc summarising all the changes and reasons for them is in progress. * fix ci failures (mainly sphinx errors) * remove unused code * fix typo * Pin pytest version to <7.2.0 * reformat code * address pr feedback * address PR feedback * address pr feedback * Update google/api_core/future/polling.py Co-authored-by: Victor Chudnovsky <vchudnov@google.com> * Apply documentation suggestions from code review Co-authored-by: Victor Chudnovsky <vchudnov@google.com> * Address PR feedback Co-authored-by: Victor Chudnovsky <vchudnov@google.com> * feat: Allow representing enums with their unqualified symbolic names in headers (#465) * feat: Allow non-fully-qualified enums in routing headers * Rename s/fully_qualified_enums/qualified_enums/g for correctness * chore: minor tweaks * chore: Temporary workaround for pytest in noxfile. * Fix import order * bring coverage to 100% * lint * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * remove replacement in owlbot.py causing lint failure Co-authored-by: Anthonios Partheniou <partheniou@google.com> Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com> * chore(python): update release script dependencies (#472) * chore(python): drop flake8-import-order in samples noxfile Source-Link: googleapis/synthtool@6ed3a83 Post-Processor: gcr.io/cloud-devrel-public-resources/owlbot-python:latest@sha256:3abfa0f1886adaf0b83f07cb117b24a639ea1cb9cffe56d43280b977033563eb * drop flake8-import-order * lint * use python 3.9 for docs * resolve mypy error * update python version for lint * fix lint * fix lint Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com> Co-authored-by: Anthonios Partheniou <partheniou@google.com> Co-authored-by: Vadym Matsishevskyi <25311427+vam-google@users.noreply.github.com> Co-authored-by: Victor Chudnovsky <vchudnov@google.com> Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com> Co-authored-by: gcf-owl-bot[bot] <78513119+gcf-owl-bot[bot]@users.noreply.github.com>

* fix: Fix timeout default values Also fix LRO for REST transport. This PR makes generated gapics appeciate timeout values from grpc_service_config.json instead of overriding them with None (which means no timeout) It is basically a direct fix for #1477. This PR depends on googleapis/python-api-core#462, and expects `setup.py.j2` templates to be updated after googleapis/python-api-core#462 gets pushed and released with new version. * rename uri_prefix to path_prefix to match corresponding python-api-core change * fix unnecessary `gapic_v1.method.DEFAULT` in rest stubs * fix(deps): require google-api-core >=1.34.0 * fix(deps): require google-api-core >=2.11.0 * revert changes to WORKSPACE * fix typo * fix mypy error * revert local change for debugging Co-authored-by: Anthonios Partheniou <partheniou@google.com>

tswast · 2022-12-14T22:14:57Z

google/api_core/future/polling.py


        try:
-            kwargs = {} if retry is DEFAULT_RETRY else {"retry": retry}


FYI: this was a breaking change. #477

Also for python-aiplatform: googleapis/python-aiplatform#1870

that rety logic line never worked (the retry had been i. What broke you is most likely the new default timeout value (instead of None).

fix(components): Limit google-api-core version to avoid timeout introduced in googleapis/python-api-core#462 PiperOrigin-RevId: 507613610

fix(components): Limit google-api-core version to avoid timeout introduced in googleapis/python-api-core#462 PiperOrigin-RevId: 508463716

jacobg · 2023-03-08T18:31:10Z

It seems like this merge broke using freezegun, particularly with google-cloud-datastore in unit testing. Here's the stack trace:

  File "/Users/jacob/.pyenv/versions/gae2/lib/python3.9/site-packages/google/cloud/datastore/transaction.py", line 231, in begin
    response_pb = self._client._datastore_api.begin_transaction(
  File "/Users/jacob/.pyenv/versions/gae2/lib/python3.9/site-packages/google/cloud/datastore_v1/services/datastore/client.py", line 834, in begin_transaction
    response = rpc(
  File "/Users/jacob/.pyenv/versions/gae2/lib/python3.9/site-packages/google/api_core/gapic_v1/method.py", line 113, in __call__
    return wrapped_func(*args, **kwargs)
  File "/Users/jacob/.pyenv/versions/gae2/lib/python3.9/site-packages/google/api_core/timeout.py", line 120, in func_with_timeout
    return func(*args, **kwargs)
  File "/Users/jacob/.pyenv/versions/gae2/lib/python3.9/site-packages/google/api_core/grpc_helpers.py", line 74, in error_remapped_callable
    raise exceptions.from_grpc_error(exc) from exc
google.api_core.exceptions.DeadlineExceeded: 504 Deadline Exceeded

Would you expect that to happen? Any workarounds?

vam-google requested review from a team as code owners October 19, 2022 02:43

product-auto-label bot added the size: xl Pull request size is extra large. label Oct 19, 2022

vam-google requested review from parthea and atulep October 19, 2022 02:43

vam-google changed the title ~~fix: Major refactoring and fix for Polling, Retry and Timeout logic~~ fix: Major refactoring of Polling, Retry and Timeout logic Oct 19, 2022

fix ci failures (mainly sphinx errors)

6a63d34

This was referenced Oct 20, 2022

Missing default deadlines in generated stubs googleapis/gapic-generator-python#1477

Closed

retry parameter ignored in the operation.result(retry=retry) call #458

Closed

igorbernstein2 mentioned this pull request Oct 20, 2022

fix: respect deadlines for column family operations googleapis/python-bigtable#687

Merged

vam-google mentioned this pull request Oct 27, 2022

fix: Fix timeout default values googleapis/gapic-generator-python#1480

Merged

atulep reviewed Oct 28, 2022

View reviewed changes

google/api_core/future/polling.py Outdated Show resolved Hide resolved

vam-google added 4 commits October 31, 2022 11:10

remove unused code

ea8b176

fix typo

ad03684

Pin pytest version to <7.2.0

eba0f38

reformat code

23c02c6