[fix][broker] Avoid block markDeletePosition forward when skip lost entries

[hanmz]

Motivation

The configuration autoSkipNonRecoverableData is designed to turn this feature on if we can accept partial data loss. When some entries is lost, the broker will still work. cursor.readPosition will move to next position for skip these positions.
But now we have this problem: cursor.readPosition can move to next positition, but the cursor.markDeletePosition can not forward.

for example:

1. topic-1 has some entries: [1:1],[1:2],[1:3],[1:4],[1:5],[1:6],[1:7],[1:8]
2. entry [1,4] and [1,5] lost
3. cursor read entry from ledger, then send message [1:1],[1:2],[1:3],[1:6],[1:7],[1:8] to consumer. (Current: readPosition=[1:8], markDeletePosition=[1:1])
4. consumer individual ack [1:1],[1:2],[1:3],[1:6],[1:7],[1:8] (Current: readPosition=[1:8], markDeletePosition=[1:3])
5. markDeletePosition will stay at [1:3], can not move forward to [1:8]

The root cause is: When entries lost, lost entries not be delete from cursor.

Modifications

When skip lost entries, delete these entries from cursor.

Documentation

• doc
• doc-required
• doc-not-needed
• doc-complete

[hanmz]

/pulsarbot rerun-failure-checks

[hanmz]

/pulsarbot rerun-failure-checks

[github-actions]

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