[DISCUSS] Deprecate Hadoop source method from (batch) ExecutionEnvironment

[Fabian Hueske-2]

Hi everybody,

I would like to propose to deprecate the utility methods to read data with Hadoop InputFormats from the (batch) ExecutionEnvironment.

The motivation for deprecating these methods is reduce Flink's dependency on Hadoop but rather have Hadoop as an optional dependency for users that actually need it (HDFS, MapRed-Compat, ...). Eventually, we want to have Flink distribution that does not have a hard Hadoop dependency.

One step for this is to remove the Hadoop dependency from flink-java (Flink's Java DataSet API) which is currently required due to the above utility methods (see FLINK-4315). We recently received a PR that addresses FLINK-4315 and removes the Hadoop methods from the ExecutionEnvironment. After some discussion, it was decided to defer the PR to Flink 2.0 because it breaks the API (these methods are delared @PublicEvolving).

I propose to accept this PR for Flink 1.2, but instead of removing the methods deprecating them.

This would help to migrate old code and prevent new usage of these methods.

For a later Flink release (1.3 or 2.0) we could remove these methods and the Hadoop dependency on flink-java.

What do others think?

Best, Fabian

[Till Rohrmann]

Fabian's proposal sounds good to me. It would be a good first step towards removing our dependency on Hadoop.

Thus, +1 for the changes.

Cheers,

Till

On Fri, Oct 14, 2016 at 11:29 AM, Fabian Hueske <[hidden email]> wrote:

Hi everybody,

I would like to propose to deprecate the utility methods to read data with Hadoop InputFormats from the (batch) ExecutionEnvironment.

The motivation for deprecating these methods is reduce Flink's dependency on Hadoop but rather have Hadoop as an optional dependency for users that actually need it (HDFS, MapRed-Compat, ...). Eventually, we want to have Flink distribution that does not have a hard Hadoop dependency.

One step for this is to remove the Hadoop dependency from flink-java (Flink's Java DataSet API) which is currently required due to the above utility methods (see FLINK-4315). We recently received a PR that addresses FLINK-4315 and removes the Hadoop methods from the ExecutionEnvironment. After some discussion, it was decided to defer the PR to Flink 2.0 because it breaks the API (these methods are delared @PublicEvolving).

I propose to accept this PR for Flink 1.2, but instead of removing the methods deprecating them.

This would help to migrate old code and prevent new usage of these methods.

For a later Flink release (1.3 or 2.0) we could remove these methods and the Hadoop dependency on flink-java.

What do others think?

Best, Fabian

[Aljoscha Krettek]

+1 for deprecating and the removing.

On Fri, 14 Oct 2016 at 11:38 Till Rohrmann <[hidden email]> wrote:

Fabian's proposal sounds good to me. It would be a good first step towards removing our dependency on Hadoop.

Thus, +1 for the changes.

Cheers,

Till

On Fri, Oct 14, 2016 at 11:29 AM, Fabian Hueske <[hidden email]> wrote:

Hi everybody,

I would like to propose to deprecate the utility methods to read data with Hadoop InputFormats from the (batch) ExecutionEnvironment.

The motivation for deprecating these methods is reduce Flink's dependency on Hadoop but rather have Hadoop as an optional dependency for users that actually need it (HDFS, MapRed-Compat, ...). Eventually, we want to have Flink distribution that does not have a hard Hadoop dependency.

One step for this is to remove the Hadoop dependency from flink-java (Flink's Java DataSet API) which is currently required due to the above utility methods (see FLINK-4315). We recently received a PR that addresses FLINK-4315 and removes the Hadoop methods from the ExecutionEnvironment. After some discussion, it was decided to defer the PR to Flink 2.0 because it breaks the API (these methods are delared @PublicEvolving).

I propose to accept this PR for Flink 1.2, but instead of removing the methods deprecating them.

This would help to migrate old code and prevent new usage of these methods.

For a later Flink release (1.3 or 2.0) we could remove these methods and the Hadoop dependency on flink-java.

What do others think?

Best, Fabian

[Shannon Carey]

Speaking as a user, if you are suggesting that you will retain the functionality but move the methods to an optional dependency, it makes sense to me. We have used the Hadoop integration for AvroParquetInputFormat and CqlBulkOutputFormat in Flink (although we won't be using CqlBulkOutputFormat any longer because it doesn't seem to be reliable).

-Shannon

From: Fabian Hueske <[hidden email]>
Date: Friday, October 14, 2016 at 4:29 AM
To: <[hidden email]>, "[hidden email]" <[hidden email]>
Subject: [DISCUSS] Deprecate Hadoop source method from (batch) ExecutionEnvironment

Hi everybody,

I would like to propose to deprecate the utility methods to read data with Hadoop InputFormats from the (batch) ExecutionEnvironment.

The motivation for deprecating these methods is reduce Flink's dependency on Hadoop but rather have Hadoop as an optional dependency for users that actually need it (HDFS, MapRed-Compat, ...). Eventually, we want to have Flink distribution that does not have a hard Hadoop dependency.

One step for this is to remove the Hadoop dependency from flink-java (Flink's Java DataSet API) which is currently required due to the above utility methods (see FLINK-4315). We recently received a PR that addresses FLINK-4315 and removes the Hadoop methods from the ExecutionEnvironment. After some discussion, it was decided to defer the PR to Flink 2.0 because it breaks the API (these methods are delared @PublicEvolving).

I propose to accept this PR for Flink 1.2, but instead of removing the methods deprecating them.

This would help to migrate old code and prevent new usage of these methods.

For a later Flink release (1.3 or 2.0) we could remove these methods and the Hadoop dependency on flink-java.

What do others think?

Best, Fabian

[Fabian Hueske-2]

Hi Shannon,

the plan is as follows:

We will keep the methods as they are for 1.2 but deprecate them and at the same time we will add alternatives in an optional dependency.

In a later release, the deprecated methods will be removed and everybody has to switch to the optional dependency.

Does that work for you?

Best, Fabian

2016-10-14 17:30 GMT+02:00 Shannon Carey <[hidden email]>:

Speaking as a user, if you are suggesting that you will retain the functionality but move the methods to an optional dependency, it makes sense to me. We have used the Hadoop integration for AvroParquetInputFormat and CqlBulkOutputFormat in Flink (although we won't be using CqlBulkOutputFormat any longer because it doesn't seem to be reliable).

-Shannon

From: Fabian Hueske <[hidden email]>
Date: Friday, October 14, 2016 at 4:29 AM
To: <[hidden email]>, "[hidden email]" <[hidden email]>
Subject: [DISCUSS] Deprecate Hadoop source method from (batch) ExecutionEnvironment

Hi everybody,

I would like to propose to deprecate the utility methods to read data with Hadoop InputFormats from the (batch) ExecutionEnvironment.

The motivation for deprecating these methods is reduce Flink's dependency on Hadoop but rather have Hadoop as an optional dependency for users that actually need it (HDFS, MapRed-Compat, ...). Eventually, we want to have Flink distribution that does not have a hard Hadoop dependency.

One step for this is to remove the Hadoop dependency from flink-java (Flink's Java DataSet API) which is currently required due to the above utility methods (see FLINK-4315). We recently received a PR that addresses FLINK-4315 and removes the Hadoop methods from the ExecutionEnvironment. After some discussion, it was decided to defer the PR to Flink 2.0 because it breaks the API (these methods are delared @PublicEvolving).

I propose to accept this PR for Flink 1.2, but instead of removing the methods deprecating them.

This would help to migrate old code and prevent new usage of these methods.

For a later Flink release (1.3 or 2.0) we could remove these methods and the Hadoop dependency on flink-java.

What do others think?

Best, Fabian

[Shannon Carey]

Yep!

From: Fabian Hueske <[hidden email]>
Date: Friday, October 14, 2016 at 11:00 AM
To: Shannon Carey <[hidden email]>
Cc: "[hidden email]" <[hidden email]>
Subject: Re: [DISCUSS] Deprecate Hadoop source method from (batch) ExecutionEnvironment

Hi Shannon,

the plan is as follows:

We will keep the methods as they are for 1.2 but deprecate them and at the same time we will add alternatives in an optional dependency.

In a later release, the deprecated methods will be removed and everybody has to switch to the optional dependency.

Does that work for you?

Best, Fabian

2016-10-14 17:30 GMT+02:00 Shannon Carey <[hidden email]>:

Speaking as a user, if you are suggesting that you will retain the functionality but move the methods to an optional dependency, it makes sense to me. We have used the Hadoop integration for AvroParquetInputFormat and CqlBulkOutputFormat in Flink (although we won't be using CqlBulkOutputFormat any longer because it doesn't seem to be reliable).

-Shannon

From: Fabian Hueske <[hidden email]>
Date: Friday, October 14, 2016 at 4:29 AM
To: <[hidden email]>, "[hidden email]" <[hidden email]>
Subject: [DISCUSS] Deprecate Hadoop source method from (batch) ExecutionEnvironment

Hi everybody,

I would like to propose to deprecate the utility methods to read data with Hadoop InputFormats from the (batch) ExecutionEnvironment.

The motivation for deprecating these methods is reduce Flink's dependency on Hadoop but rather have Hadoop as an optional dependency for users that actually need it (HDFS, MapRed-Compat, ...). Eventually, we want to have Flink distribution that does not have a hard Hadoop dependency.

One step for this is to remove the Hadoop dependency from flink-java (Flink's Java DataSet API) which is currently required due to the above utility methods (see FLINK-4315). We recently received a PR that addresses FLINK-4315 and removes the Hadoop methods from the ExecutionEnvironment. After some discussion, it was decided to defer the PR to Flink 2.0 because it breaks the API (these methods are delared @PublicEvolving).

I propose to accept this PR for Flink 1.2, but instead of removing the methods deprecating them.

This would help to migrate old code and prevent new usage of these methods.

For a later Flink release (1.3 or 2.0) we could remove these methods and the Hadoop dependency on flink-java.

What do others think?

Best, Fabian