Blocks need to be restored from temporary table upon failure or application restart

[michielmulders]

Expected behavior

Whenever the node crashes or upon failure during one of the sync processes we roll back to the original state by applying the blocks in the temp table and continue the operating of the node.

Functions to be implemented:

• restoreBlocks()

1. Read blocks from temp table
2. Apply blocks (blocks verify process - do not care about how it is implemented - just call this)
3. Remove temp blocks after the apply process is successful

Design For Issue

Place the restoreBlocks in blocks/blocks.
The synchronizer class should injects blocks module as a dependency so we can access restoreBlocks inside.

Which version(s) does this affect? (Environment, OS, etc...)

feature/improve-bft

Tasks list

• Add functionality for applying block from blocks temporary table.
• Restore blocks from the temporary table when the node restarts.
• Restore blocks from the temporary table when the synchronization fails.
• Write integration tests.

[sridharmeganathan]

@michielmulders will do the analysis before the next sprint grooming meeting.

[michielmulders]

Discusssion conclusion about the approach of storing temporary data.

[michielmulders]

First, happy path -> block in temp table and restore to chain (via test)