Handle exceptions from ingestion server's calls to Elasticsearch

[dhruvkb]

Description

Handle exceptions from ingestion server's calls to Elasticsearch.

Example

The following calls to Elasticsearch to create a new index can raise an exception, if that index already exists.

BadRequestError(400, 'resource_already_exists_exception', 'index [image-init-filtered/-jGDxWxbQS-_SdarNzk34g] already exists')

openverse/ingestion_server/ingestion_server/indexer.py

Lines 302 to 305 in 46b6b93

	self.es.indices.create(
	index=destination_index,
	body=index_settings(model_name),
	)

openverse/ingestion_server/ingestion_server/indexer.py

Lines 499 to 502 in 46b6b93

	self.es.indices.create(
	index=destination_index,
	body=index_settings(model_name),
	)

This exception is not handled in the code. It should be captured and should mark the appropriate task as 'succeeded' in this particular case because the task is idempotent and the index has been created.

Additional context

This exception was captured by Sentry.

[sentry-io]

Sentry issue: OPENVERSE-INGESTION_SERVER-12

[sentry-io]

Sentry issue: OPENVERSE-INGESTION_SERVER-10

[AetherUnbound]

After looking into this, it looks like all of the events related to this error are in the local environment. Additionally, I feel like that piece of the code should fail the way it's currently set up to do in production; the attempted recreation of an existing index, while idempotent at that particular step, would likely mean a failure elsewhere in our process. It'd be best for execution to stop here in that case I think!

I'm going to close this issue for now.