yihua commented on code in PR #5113: URL: https://github.com/apache/hudi/pull/5113#discussion_r984024879
########## rfc/rfc-56/rfc-56.md: ########## @@ -0,0 +1,226 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> + +# RFC-56: Federated Storage Layer + +## Proposers +- @umehrot2 + +## Approvers +- @vinoth +- @shivnarayan + +## Status + +JIRA: [https://issues.apache.org/jira/browse/HUDI-3625](https://issues.apache.org/jira/browse/HUDI-3625) + +## Abstract + +As you scale your Apache Hudi workloads over Cloud object stores like Amazon S3, there is potential of hitting request +throttling limits which in-turn impacts performance. In this RFC, we are proposing to support an alternate storage +layout that is optimized for Amazon S3 and other cloud object stores, which helps achieve maximum throughput and +significantly reduce throttling. + +In addition, we are proposing an interface that would allow users to implement their own custom strategy to allow them +to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases. + +## Background + +Apache Hudi follows the traditional Hive storage layout while writing files on storage: +- Partitioned Tables: The files are distributed across multiple physical partition folders, under the table's base path. +- Non Partitioned Tables: The files are stored directly under the table's base path. + +While this storage layout scales well for HDFS, it increases the probability of hitting request throttle limits when +working with cloud object stores like Amazon S3 and others. This is because Amazon S3 and other cloud stores [throttle +requests based on object prefix](https://aws.amazon.com/premiumsupport/knowledge-center/s3-request-limit-avoid-throttling/). +Amazon S3 does scale based on request patterns for different prefixes and adds internal partitions (with their own request limits), +but there can be a 30 - 60 minute wait time before new partitions are created. Thus, all files/objects stored under the +same table path prefix could result in these request limits being hit for the table prefix, specially as workloads +scale, and there are several thousands of files being written/updated concurrently. This hurts performance due to +re-trying of failed requests affecting throughput, and result in occasional failures if the retries are not able to +succeed either and continue to be throttled. + +The traditional storage layout also tightly couples the partitions as folders under the table path. However, +some users want flexibility to be able to distribute files/partitions under multiple different paths across cloud stores, +hdfs etc. based on their specific needs. For example, customers have use cases to distribute files for each partition under +a separate S3 bucket with its individual encryption key. It is not possible to implement such use-cases with Hudi currently. + +The high level proposal here is to introduce a new storage layout strategy, where all files are distributed evenly across +multiple randomly generated prefixes under the Amazon S3 bucket, instead of being stored under a common table path/prefix. +This would help distribute the requests evenly across different prefixes, resulting in Amazon S3 to create partitions for +the prefixes each with its own request limit. This significantly reduces the possibility of hitting the request limit +for a specific prefix/partition. + +In addition, we want to expose an interface that provides users the flexibility to implement their own strategy for +distributing files if using the traditional Hive storage layout or federated storage layer (proposed in this RFC) does +not meet their use-case. + +## Design + +### Interface + +```java +/** + * Interface for providing storage file locations. + */ +public interface FederatedStorageStrategy extends Serializable { + /** + * Return a fully-qualified storage file location for the given filename. + * + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String fileName); + + /** + * Return a fully-qualified storage file location for the given partition and filename. + * + * @param partitionPath partition path for the file + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String partitionPath, String fileName); Review Comment: What does the `fileName` refer to here? Is it the logical file name of a base or log file in a Hudi file slice? And is this relative or absolute? ########## rfc/rfc-56/rfc-56.md: ########## @@ -0,0 +1,226 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> + +# RFC-56: Federated Storage Layer + +## Proposers +- @umehrot2 + +## Approvers +- @vinoth +- @shivnarayan + +## Status + +JIRA: [https://issues.apache.org/jira/browse/HUDI-3625](https://issues.apache.org/jira/browse/HUDI-3625) + +## Abstract + +As you scale your Apache Hudi workloads over Cloud object stores like Amazon S3, there is potential of hitting request +throttling limits which in-turn impacts performance. In this RFC, we are proposing to support an alternate storage +layout that is optimized for Amazon S3 and other cloud object stores, which helps achieve maximum throughput and +significantly reduce throttling. + +In addition, we are proposing an interface that would allow users to implement their own custom strategy to allow them +to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases. Review Comment: > allow them to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases Does this mean that for a single Hudi table, some partitions and data files are on one cloud, while others are on a different cloud or availability zone? What's the use case of doing this? Now we always assume that the single Hudi table resides on the same File System. ########## rfc/rfc-56/rfc-56.md: ########## @@ -0,0 +1,226 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> + +# RFC-56: Federated Storage Layer + +## Proposers +- @umehrot2 + +## Approvers +- @vinoth +- @shivnarayan + +## Status + +JIRA: [https://issues.apache.org/jira/browse/HUDI-3625](https://issues.apache.org/jira/browse/HUDI-3625) + +## Abstract + +As you scale your Apache Hudi workloads over Cloud object stores like Amazon S3, there is potential of hitting request +throttling limits which in-turn impacts performance. In this RFC, we are proposing to support an alternate storage +layout that is optimized for Amazon S3 and other cloud object stores, which helps achieve maximum throughput and +significantly reduce throttling. + +In addition, we are proposing an interface that would allow users to implement their own custom strategy to allow them +to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases. + +## Background + +Apache Hudi follows the traditional Hive storage layout while writing files on storage: +- Partitioned Tables: The files are distributed across multiple physical partition folders, under the table's base path. +- Non Partitioned Tables: The files are stored directly under the table's base path. + +While this storage layout scales well for HDFS, it increases the probability of hitting request throttle limits when +working with cloud object stores like Amazon S3 and others. This is because Amazon S3 and other cloud stores [throttle +requests based on object prefix](https://aws.amazon.com/premiumsupport/knowledge-center/s3-request-limit-avoid-throttling/). +Amazon S3 does scale based on request patterns for different prefixes and adds internal partitions (with their own request limits), +but there can be a 30 - 60 minute wait time before new partitions are created. Thus, all files/objects stored under the +same table path prefix could result in these request limits being hit for the table prefix, specially as workloads +scale, and there are several thousands of files being written/updated concurrently. This hurts performance due to +re-trying of failed requests affecting throughput, and result in occasional failures if the retries are not able to +succeed either and continue to be throttled. + +The traditional storage layout also tightly couples the partitions as folders under the table path. However, +some users want flexibility to be able to distribute files/partitions under multiple different paths across cloud stores, +hdfs etc. based on their specific needs. For example, customers have use cases to distribute files for each partition under +a separate S3 bucket with its individual encryption key. It is not possible to implement such use-cases with Hudi currently. + +The high level proposal here is to introduce a new storage layout strategy, where all files are distributed evenly across +multiple randomly generated prefixes under the Amazon S3 bucket, instead of being stored under a common table path/prefix. +This would help distribute the requests evenly across different prefixes, resulting in Amazon S3 to create partitions for +the prefixes each with its own request limit. This significantly reduces the possibility of hitting the request limit +for a specific prefix/partition. + +In addition, we want to expose an interface that provides users the flexibility to implement their own strategy for +distributing files if using the traditional Hive storage layout or federated storage layer (proposed in this RFC) does +not meet their use-case. + +## Design + +### Interface + +```java +/** + * Interface for providing storage file locations. + */ +public interface FederatedStorageStrategy extends Serializable { + /** + * Return a fully-qualified storage file location for the given filename. + * + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String fileName); + + /** + * Return a fully-qualified storage file location for the given partition and filename. + * + * @param partitionPath partition path for the file + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String partitionPath, String fileName); +} +``` + +### Generating file paths for Cloud storage optimized layout + +We want to distribute files evenly across multiple random prefixes, instead of following the traditional Hive storage +layout of keeping them under a common table path/prefix. In addition to the `Table Path`, for this new layout user will +configure another `Table Storage Path` under which the actual data files will be distributed. The original `Table Path` will +be used to maintain the table/partitions Hudi metadata. + +For the purpose of this documentation lets assume: +``` +Table Path => s3://<table_bucket>/<hudi_table_name>/ + +Table Storage Path => s3://<table_storage_bucket>/ +``` +Note: `Table Storage Path` can be a path in the same Amazon S3 bucket or a different bucket. For best results, +`Table Storage Path` should be a bucket instead of a prefix under the bucket as it allows for S3 to partition sooner. + +We will use a Hashing function on the `File Name` to map them to a prefix generated under `Table Storage Path`: +``` +s3://<table_storage_bucket>/<hash_prefix>/.. +``` + +In addition, under the hash prefix we will follow a folder structure by appending Hudi Table Name and Partition. This +folder structuring would be useful if we ever have to do a file system listing to re-create the metadata file list for +the table (discussed more in the next section). Here is how the final layout would look like for `partitioned` tables: +``` +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=usa/075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=india/4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=india/.9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.log.1_0-22-26 +... +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=india/9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=uk/a62aa56b-d55e-4a2b-88a6-d603ef26775c-0_8-11-50_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=india/.4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.log.1_0-22-26 +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=usa/.075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.log.1_0-22-26 +... +``` +For `non-partitioned` tables, this is how it would look: +``` +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/.9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.log.1_0-22-26 +... +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/a62aa56b-d55e-4a2b-88a6-d603ef26775c-0_8-11-50_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/.4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.log.1_0-22-26 +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/.075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.log.1_0-22-26 +... +``` +**Note**: For `Merge on Read` tables, the log files will also go through the same hashing process and may not end up under Review Comment: Basically, the merging logic takes the logical file names based on the file group, and resolves the actual file names on storage based on the storage layout. Then it should still work. ########## rfc/rfc-56/rfc-56.md: ########## @@ -0,0 +1,226 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> + +# RFC-56: Federated Storage Layer + +## Proposers +- @umehrot2 + +## Approvers +- @vinoth +- @shivnarayan + +## Status + +JIRA: [https://issues.apache.org/jira/browse/HUDI-3625](https://issues.apache.org/jira/browse/HUDI-3625) + +## Abstract + +As you scale your Apache Hudi workloads over Cloud object stores like Amazon S3, there is potential of hitting request +throttling limits which in-turn impacts performance. In this RFC, we are proposing to support an alternate storage +layout that is optimized for Amazon S3 and other cloud object stores, which helps achieve maximum throughput and +significantly reduce throttling. + +In addition, we are proposing an interface that would allow users to implement their own custom strategy to allow them +to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases. + +## Background + +Apache Hudi follows the traditional Hive storage layout while writing files on storage: +- Partitioned Tables: The files are distributed across multiple physical partition folders, under the table's base path. +- Non Partitioned Tables: The files are stored directly under the table's base path. + +While this storage layout scales well for HDFS, it increases the probability of hitting request throttle limits when +working with cloud object stores like Amazon S3 and others. This is because Amazon S3 and other cloud stores [throttle +requests based on object prefix](https://aws.amazon.com/premiumsupport/knowledge-center/s3-request-limit-avoid-throttling/). +Amazon S3 does scale based on request patterns for different prefixes and adds internal partitions (with their own request limits), +but there can be a 30 - 60 minute wait time before new partitions are created. Thus, all files/objects stored under the +same table path prefix could result in these request limits being hit for the table prefix, specially as workloads +scale, and there are several thousands of files being written/updated concurrently. This hurts performance due to +re-trying of failed requests affecting throughput, and result in occasional failures if the retries are not able to +succeed either and continue to be throttled. + +The traditional storage layout also tightly couples the partitions as folders under the table path. However, +some users want flexibility to be able to distribute files/partitions under multiple different paths across cloud stores, +hdfs etc. based on their specific needs. For example, customers have use cases to distribute files for each partition under +a separate S3 bucket with its individual encryption key. It is not possible to implement such use-cases with Hudi currently. + +The high level proposal here is to introduce a new storage layout strategy, where all files are distributed evenly across +multiple randomly generated prefixes under the Amazon S3 bucket, instead of being stored under a common table path/prefix. Review Comment: Adding to what @alexeykudinkin mentions, for the "common prefix" problem, the logical partitioning solves the listing performance and partitioning, but during the writing, if all data files are under the same table path without physical partitioning, writing data files (PUT requests) may still get bottlenecked on the throttling as they have the same prefix. So randomly generated prefixes can still help here. ########## rfc/rfc-56/rfc-56.md: ########## @@ -0,0 +1,226 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> + +# RFC-56: Federated Storage Layer + +## Proposers +- @umehrot2 + +## Approvers +- @vinoth +- @shivnarayan + +## Status + +JIRA: [https://issues.apache.org/jira/browse/HUDI-3625](https://issues.apache.org/jira/browse/HUDI-3625) + +## Abstract + +As you scale your Apache Hudi workloads over Cloud object stores like Amazon S3, there is potential of hitting request +throttling limits which in-turn impacts performance. In this RFC, we are proposing to support an alternate storage +layout that is optimized for Amazon S3 and other cloud object stores, which helps achieve maximum throughput and +significantly reduce throttling. + +In addition, we are proposing an interface that would allow users to implement their own custom strategy to allow them +to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases. + +## Background + +Apache Hudi follows the traditional Hive storage layout while writing files on storage: +- Partitioned Tables: The files are distributed across multiple physical partition folders, under the table's base path. +- Non Partitioned Tables: The files are stored directly under the table's base path. + +While this storage layout scales well for HDFS, it increases the probability of hitting request throttle limits when +working with cloud object stores like Amazon S3 and others. This is because Amazon S3 and other cloud stores [throttle +requests based on object prefix](https://aws.amazon.com/premiumsupport/knowledge-center/s3-request-limit-avoid-throttling/). +Amazon S3 does scale based on request patterns for different prefixes and adds internal partitions (with their own request limits), +but there can be a 30 - 60 minute wait time before new partitions are created. Thus, all files/objects stored under the +same table path prefix could result in these request limits being hit for the table prefix, specially as workloads +scale, and there are several thousands of files being written/updated concurrently. This hurts performance due to +re-trying of failed requests affecting throughput, and result in occasional failures if the retries are not able to +succeed either and continue to be throttled. + +The traditional storage layout also tightly couples the partitions as folders under the table path. However, +some users want flexibility to be able to distribute files/partitions under multiple different paths across cloud stores, +hdfs etc. based on their specific needs. For example, customers have use cases to distribute files for each partition under +a separate S3 bucket with its individual encryption key. It is not possible to implement such use-cases with Hudi currently. + +The high level proposal here is to introduce a new storage layout strategy, where all files are distributed evenly across +multiple randomly generated prefixes under the Amazon S3 bucket, instead of being stored under a common table path/prefix. +This would help distribute the requests evenly across different prefixes, resulting in Amazon S3 to create partitions for +the prefixes each with its own request limit. This significantly reduces the possibility of hitting the request limit +for a specific prefix/partition. + +In addition, we want to expose an interface that provides users the flexibility to implement their own strategy for +distributing files if using the traditional Hive storage layout or federated storage layer (proposed in this RFC) does +not meet their use-case. + +## Design + +### Interface + +```java +/** + * Interface for providing storage file locations. + */ +public interface FederatedStorageStrategy extends Serializable { + /** + * Return a fully-qualified storage file location for the given filename. + * + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String fileName); + + /** + * Return a fully-qualified storage file location for the given partition and filename. + * + * @param partitionPath partition path for the file + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String partitionPath, String fileName); +} +``` + +### Generating file paths for Cloud storage optimized layout + +We want to distribute files evenly across multiple random prefixes, instead of following the traditional Hive storage +layout of keeping them under a common table path/prefix. In addition to the `Table Path`, for this new layout user will +configure another `Table Storage Path` under which the actual data files will be distributed. The original `Table Path` will +be used to maintain the table/partitions Hudi metadata. + +For the purpose of this documentation lets assume: +``` +Table Path => s3://<table_bucket>/<hudi_table_name>/ + +Table Storage Path => s3://<table_storage_bucket>/ +``` +Note: `Table Storage Path` can be a path in the same Amazon S3 bucket or a different bucket. For best results, +`Table Storage Path` should be a bucket instead of a prefix under the bucket as it allows for S3 to partition sooner. + +We will use a Hashing function on the `File Name` to map them to a prefix generated under `Table Storage Path`: +``` +s3://<table_storage_bucket>/<hash_prefix>/.. +``` + +In addition, under the hash prefix we will follow a folder structure by appending Hudi Table Name and Partition. This +folder structuring would be useful if we ever have to do a file system listing to re-create the metadata file list for +the table (discussed more in the next section). Here is how the final layout would look like for `partitioned` tables: +``` +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=usa/075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=india/4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=india/.9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.log.1_0-22-26 +... +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=india/9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=uk/a62aa56b-d55e-4a2b-88a6-d603ef26775c-0_8-11-50_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=india/.4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.log.1_0-22-26 +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=usa/.075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.log.1_0-22-26 +... +``` +For `non-partitioned` tables, this is how it would look: +``` +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/.9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.log.1_0-22-26 +... +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/a62aa56b-d55e-4a2b-88a6-d603ef26775c-0_8-11-50_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/.4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.log.1_0-22-26 +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/.075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.log.1_0-22-26 +... +``` +**Note**: For `Merge on Read` tables, the log files will also go through the same hashing process and may not end up under +the same prefix as the base parquet file for the FileSlice to which it belongs. + +The original table path will continue to store the `metadata folder` and `partition metadata` files: +``` +s3://<table_bucket>/<hudi_table_name>/.hoodie/... Review Comment: Do we persist the table layout strategy in the table properties (`.hoodie/hoodie.properties`)? The table layout strategy should not change. That can also differentiate the table from Hive storage layout and avoid confusion. ########## rfc/rfc-56/rfc-56.md: ########## @@ -0,0 +1,226 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> + +# RFC-56: Federated Storage Layer + +## Proposers +- @umehrot2 + +## Approvers +- @vinoth +- @shivnarayan + +## Status + +JIRA: [https://issues.apache.org/jira/browse/HUDI-3625](https://issues.apache.org/jira/browse/HUDI-3625) + +## Abstract + +As you scale your Apache Hudi workloads over Cloud object stores like Amazon S3, there is potential of hitting request +throttling limits which in-turn impacts performance. In this RFC, we are proposing to support an alternate storage +layout that is optimized for Amazon S3 and other cloud object stores, which helps achieve maximum throughput and +significantly reduce throttling. + +In addition, we are proposing an interface that would allow users to implement their own custom strategy to allow them +to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases. + +## Background + +Apache Hudi follows the traditional Hive storage layout while writing files on storage: +- Partitioned Tables: The files are distributed across multiple physical partition folders, under the table's base path. +- Non Partitioned Tables: The files are stored directly under the table's base path. + +While this storage layout scales well for HDFS, it increases the probability of hitting request throttle limits when +working with cloud object stores like Amazon S3 and others. This is because Amazon S3 and other cloud stores [throttle +requests based on object prefix](https://aws.amazon.com/premiumsupport/knowledge-center/s3-request-limit-avoid-throttling/). +Amazon S3 does scale based on request patterns for different prefixes and adds internal partitions (with their own request limits), +but there can be a 30 - 60 minute wait time before new partitions are created. Thus, all files/objects stored under the +same table path prefix could result in these request limits being hit for the table prefix, specially as workloads +scale, and there are several thousands of files being written/updated concurrently. This hurts performance due to +re-trying of failed requests affecting throughput, and result in occasional failures if the retries are not able to +succeed either and continue to be throttled. + +The traditional storage layout also tightly couples the partitions as folders under the table path. However, +some users want flexibility to be able to distribute files/partitions under multiple different paths across cloud stores, +hdfs etc. based on their specific needs. For example, customers have use cases to distribute files for each partition under +a separate S3 bucket with its individual encryption key. It is not possible to implement such use-cases with Hudi currently. + +The high level proposal here is to introduce a new storage layout strategy, where all files are distributed evenly across +multiple randomly generated prefixes under the Amazon S3 bucket, instead of being stored under a common table path/prefix. +This would help distribute the requests evenly across different prefixes, resulting in Amazon S3 to create partitions for +the prefixes each with its own request limit. This significantly reduces the possibility of hitting the request limit +for a specific prefix/partition. + +In addition, we want to expose an interface that provides users the flexibility to implement their own strategy for +distributing files if using the traditional Hive storage layout or federated storage layer (proposed in this RFC) does +not meet their use-case. + +## Design + +### Interface + +```java +/** + * Interface for providing storage file locations. + */ +public interface FederatedStorageStrategy extends Serializable { + /** + * Return a fully-qualified storage file location for the given filename. + * + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String fileName); + + /** + * Return a fully-qualified storage file location for the given partition and filename. + * + * @param partitionPath partition path for the file + * @param fileName data file name + * @return a fully-qualified location URI for a data file + */ + String storageLocation(String partitionPath, String fileName); +} +``` + +### Generating file paths for Cloud storage optimized layout + +We want to distribute files evenly across multiple random prefixes, instead of following the traditional Hive storage +layout of keeping them under a common table path/prefix. In addition to the `Table Path`, for this new layout user will +configure another `Table Storage Path` under which the actual data files will be distributed. The original `Table Path` will +be used to maintain the table/partitions Hudi metadata. + +For the purpose of this documentation lets assume: +``` +Table Path => s3://<table_bucket>/<hudi_table_name>/ + +Table Storage Path => s3://<table_storage_bucket>/ +``` +Note: `Table Storage Path` can be a path in the same Amazon S3 bucket or a different bucket. For best results, +`Table Storage Path` should be a bucket instead of a prefix under the bucket as it allows for S3 to partition sooner. + +We will use a Hashing function on the `File Name` to map them to a prefix generated under `Table Storage Path`: +``` +s3://<table_storage_bucket>/<hash_prefix>/.. +``` + +In addition, under the hash prefix we will follow a folder structure by appending Hudi Table Name and Partition. This +folder structuring would be useful if we ever have to do a file system listing to re-create the metadata file list for +the table (discussed more in the next section). Here is how the final layout would look like for `partitioned` tables: +``` +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=usa/075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=india/4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/country=india/.9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.log.1_0-22-26 +... +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=india/9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=uk/a62aa56b-d55e-4a2b-88a6-d603ef26775c-0_8-11-50_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=india/.4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.log.1_0-22-26 +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/country=usa/.075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.log.1_0-22-26 +... +``` +For `non-partitioned` tables, this is how it would look: +``` +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.parquet +s3://<table_storage_bucket>/01f50736/<hudi_table_name>/.9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.log.1_0-22-26 +... +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/9320889c-8537-4aa7-a63e-ef088b9a21ce-0_9-11-51_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/a62aa56b-d55e-4a2b-88a6-d603ef26775c-0_8-11-50_20220301005056692.parquet +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/.4b0c6b40-2ac0-4a1c-a26f-6338aa4db22e-0_6-11-48_20220301005056692.log.1_0-22-26 +s3://<table_storage_bucket>/0bfb3d6e/<hudi_table_name>/.075f3295-def8-4a42-a927-07fd2dd2976c-0_7-11-49_20220301005056692.log.1_0-22-26 +... +``` +**Note**: For `Merge on Read` tables, the log files will also go through the same hashing process and may not end up under Review Comment: For this particular prefix-based storage layout strategy, we can do hashing based on the file group, but we can still leave the interface to support other file-level distributions. And the MOR log merging logic should not depend on the storage layout, right? So no need to couple the file group layout with the storage layout to maintain the layering? ########## rfc/rfc-56/rfc-56.md: ########## @@ -0,0 +1,226 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You under the Apache License, Version 2.0 + (the "License"); you may not use this file except in compliance with + the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. +--> + +# RFC-56: Federated Storage Layer + +## Proposers +- @umehrot2 + +## Approvers +- @vinoth +- @shivnarayan + +## Status + +JIRA: [https://issues.apache.org/jira/browse/HUDI-3625](https://issues.apache.org/jira/browse/HUDI-3625) + +## Abstract + +As you scale your Apache Hudi workloads over Cloud object stores like Amazon S3, there is potential of hitting request +throttling limits which in-turn impacts performance. In this RFC, we are proposing to support an alternate storage +layout that is optimized for Amazon S3 and other cloud object stores, which helps achieve maximum throughput and +significantly reduce throttling. + +In addition, we are proposing an interface that would allow users to implement their own custom strategy to allow them +to distribute the data files across cloud stores, hdfs or on prem based on their specific use-cases. + +## Background + +Apache Hudi follows the traditional Hive storage layout while writing files on storage: +- Partitioned Tables: The files are distributed across multiple physical partition folders, under the table's base path. +- Non Partitioned Tables: The files are stored directly under the table's base path. + +While this storage layout scales well for HDFS, it increases the probability of hitting request throttle limits when +working with cloud object stores like Amazon S3 and others. This is because Amazon S3 and other cloud stores [throttle +requests based on object prefix](https://aws.amazon.com/premiumsupport/knowledge-center/s3-request-limit-avoid-throttling/). +Amazon S3 does scale based on request patterns for different prefixes and adds internal partitions (with their own request limits), +but there can be a 30 - 60 minute wait time before new partitions are created. Thus, all files/objects stored under the +same table path prefix could result in these request limits being hit for the table prefix, specially as workloads +scale, and there are several thousands of files being written/updated concurrently. This hurts performance due to +re-trying of failed requests affecting throughput, and result in occasional failures if the retries are not able to +succeed either and continue to be throttled. + +The traditional storage layout also tightly couples the partitions as folders under the table path. However, +some users want flexibility to be able to distribute files/partitions under multiple different paths across cloud stores, +hdfs etc. based on their specific needs. For example, customers have use cases to distribute files for each partition under +a separate S3 bucket with its individual encryption key. It is not possible to implement such use-cases with Hudi currently. + +The high level proposal here is to introduce a new storage layout strategy, where all files are distributed evenly across +multiple randomly generated prefixes under the Amazon S3 bucket, instead of being stored under a common table path/prefix. +This would help distribute the requests evenly across different prefixes, resulting in Amazon S3 to create partitions for +the prefixes each with its own request limit. This significantly reduces the possibility of hitting the request limit +for a specific prefix/partition. Review Comment: +1 on dynamic prefix-splitting or rebalancing -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
