…a remote repository

Because this step is not crucial and we do not want to be overwhelmed with
unpredictable error to metric store (influx db in this case)

…ngConfig to resolve circular dependency

)

* Fix tracing to continue from existing trace created by grpc client

* Update python SDK with updated protos in Feast 0.1.x
Removal of storage specs

* Add __init__ file in feast/types to fix module not found error

* Add support to configure import job metrics in Feast Core

* Fix CoreApplicationWithNoWarehouseTest

* Update charts
Add import job metrics configuration
Update jaeger tracing configuration for serving

* Add lazy annotation in getBigQueryTrainingDatasetTemplater in TrainingConfig to resolve circular dependency

* Use provided values for INGESTION_METRICS_ENABLED

* Add Python mypy definition

* Increase version of google-cloud-storage in python sdk

* Update docs for influx_db_url field in ImportJobSpecs proto

* Write feature metrics in fixed window for better performance

* Update default image in Helm chart

* Write feature metrics only if the job type is streaming

* Set transform name for writing feature metrics to Influx DB

* Update chart, making resources limit optional
Following how stable postgresql chart specifies it
https://github.com/helm/charts/blob/master/stable/postgresql/values.yaml

* Fix the check for unbounded sources
Check against the ImportSpec.Type value (kafka or pubsub) instead of PipelineOptions.Streaming (the streaming status cannot be reliably determined at pipeline creation stage)

* Fix typo

* Update default Feast image tag used in the chart

Remove redis transaction