Fix bug with 404 on GET request for batch_request (#356)

* add a retry wrapper around the GET request * make mypy happy * add intermediate value --------- Co-authored-by: Matic Lubej <[email protected]> Co-authored-by: Ziga Luksic <[email protected]>
sentinel-hub · Nov 6, 2024 · e99aef1 · e99aef1
1 parent 2321e6b
commit e99aef1
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 3 deletions.
diff --git a/eogrow/pipelines/download_batch.py b/eogrow/pipelines/download_batch.py
@@ -49,7 +49,7 @@
 def _retry_on_404(func: Callable[P, T]) -> Callable[P, T]:
     @wraps(func)
     def retrying_func(*args: P.args, **kwargs: P.kwargs) -> T:
-        for wait_time in [0, 10, 100]:
+        for wait_time in [0, 10, 20, 100]:
             time.sleep(wait_time)  # if we start monitoring too soon we might hit a 404
             try:
                 return func(*args, **kwargs)
@@ -178,6 +178,7 @@ def run_procedure(self) -> tuple[list[str], list[str]]:
         batch_request = self._create_or_collect_batch_request()
 
         user_action = self._trigger_user_action(batch_request)
+        self._wait_for_sh_db_sync(batch_request)
 
         if user_action is BatchUserAction.ANALYSE or (
             user_action is BatchUserAction.START and batch_request.status is not BatchRequestStatus.ANALYSIS_DONE
@@ -308,6 +309,11 @@ def _trigger_user_action(self, batch_request: BatchRequest) -> BatchUserAction:
         LOGGER.info("Didn't trigger batch job because current batch request status is %s", status)
         return BatchUserAction.NONE
 
+    @_retry_on_404
+    def _wait_for_sh_db_sync(self, batch_request: BatchRequest) -> None:
+        """Wait for SH read/write databases to sync."""
+        self.batch_client.get_request(batch_request)
+
     def cache_batch_area_manager_grid(self, request_id: str) -> None:
         """This method ensures that area manager caches batch grid into the storage."""
         if self.area_manager.config.batch_id and self.area_manager.config.batch_id != request_id:

diff --git a/eogrow/utils/pipeline_chain.py b/eogrow/utils/pipeline_chain.py
@@ -40,8 +40,8 @@ def validate_pipeline_chain(pipeline_chain: list[RawConfig]) -> None:
 def run_pipeline_chain(pipeline_chain: list[RawConfig]) -> None:
     for run_config in pipeline_chain:
         run_schema = PipelineRunSchema.parse_obj(run_config)
-        runner = _pipeline_runner.options(**run_schema.pipeline_resources)  # type: ignore[attr-defined]
-        ray.get(runner.remote(run_schema.pipeline_config))
+        runner = _pipeline_runner.options(**run_schema.pipeline_resources)
+        ray.get(runner.remote(run_schema.pipeline_config))  # type: ignore [arg-type]
 
 
 @ray.remote(max_retries=0)