[feature][pulsar-io-mongo] Add support for full message synchronization

[shink]

Motivation

Now, the MongoDB source connector only supports the incremental message synchronization.
This PR adds support for full message synchronization.

Since MongDB 4.0, we can set the starting point for the change stream by the startAtOperationTime field.
So, we can set it to 0 to make start point the earliest.
See https://www.mongodb.com/docs/v4.2/reference/method/db.collection.watch/ for more information.

Modifications

1. Improve config object.
   There are some commonalities and differences between sink configuration and source configuration.
   So, I created an abstract class called MongoAbstractConfig which contains the commonalities between them.
   MongoSourceConfig and MongoSinkConfig contain the unique configuration.

2. Add support for full message synchronization in the source connector.

if (mongoSourceConfig.getSyncType() == SyncType.FULL_SYNC) {
    // sync currently existing messages
    // startAtOperationTime is the starting point for the change stream
    // setting startAtOperationTime to 0 means the start point is the earliest
    // see https://www.mongodb.com/docs/v4.2/reference/method/db.collection.watch/ for more information
    stream.startAtOperationTime(new BsonTimestamp(0L));
}

Verifying this change

• Make sure that the change passes the CI checks.

Does this pull request potentially affect one of the following parts:

No.

Documentation

Check the box below or label this PR directly.

Need to update docs?

• doc-required
  (Your PR needs to update docs and you will update later)

• doc-not-needed
  (Please explain why)

• doc
  (Your PR contains doc changes)

• doc-complete
  (Docs have been already added)

[shink]

Here is the key.

[shink]

I will finish UT later. Convert this PR to a draft temporarily.

[shink]

@315157973 PTAL 🎉

[315157973]

Please add some doc and make sure that the change passes the CI checks.

[shink]

@315157973 OK. And where should I add the doc to?

[github-actions]

The pr had no activity for 30 days, mark with Stale label.

[codelipenghui]

/pulsarbot run-failure-checks

[shink]

@codelipenghui Thank you very much! But why are there three failed checks after a rerun, and before there were four. And I didn't make changes to broker.

[shink]

@codelipenghui Hi, it seems that all the failed checks are flask tests, what should I do? Thank you!

[eolivelli]

/pulsar-bot rerun-failure-checks

[poorbarcode]

/pulsarbot rerun-failure-checks

[shink]

/pulsarbot rerun-failure-checks

[poorbarcode]

Hi @shink

https://github.com/apache/pulsar/actions/runs/3071925540/jobs/4963093961

this is a flaky test, It should be fine if it is run several times

https://github.com/apache/pulsar/actions/runs/3071925517/jobs/4963032830

I read the log and found that instability was caused by mockito. It should be fine if it is run several times

Many tests in this group(Pulsar CI Flaky) are unstable, but this does not affect merge. These problems are being fixed

[shink]

@poorbarcode Thank you so much for your help!

[shink]

/pulsarbot rerun-failure-checks

[shink]

/pulsarbot rerun-failure-checks

[momo-jun]

@shink seems you've already included docs in this PR, so I will update the doc-required label to doc. Thank you for adding the docs!