Feature/#2429 fork id improvement

[ARR552]

Closes #2429

What does this PR do?

This Pr adds a forkID table. If a component needs the forkID list goes to db instead of to read the info to L1. If the table is empty the synchronizer fills it until the current synced batch. During the sync process, new forkIDs are stored in db and loaded in memory.
*Important points: *
If the synchronization has not finished, do not start the aggregator because the currentForkID could be wrong.
If the node is running using microservices, other components needs to be restarted in order to be aware of the new forkID. The synchronizer is the only one that will realize that the forkID has changed. If all the components are running together, this is not a problem because the synchronizer will keep updated the memory variables shared by all the components.

Config changes:
[Etherman]
ForkIDChunkSize = 20000

Reviewers

@agnusmor
@ToniRamirezM
@joanestebanr

[agnusmor]

We are not storing this initial forkID in the db

[ARR552]

Yes, we are. Why do you say we are not storing it?

[agnusmor]

I think we can change this if/else block code for this one, since the if and else code are the same, since I understand also that we need to store the forkid interval in the case of the initial fork (genesis):

lastBlockNumber := genesisBlockNumber
if lastBlock != nil {
   lastBlockNumber := lastBlock.BlockNumber
}

log.Info("Getting forkIDs intervals. Please wait...")
// Read Fork ID FROM POE SC
forkIntervals, err := etherman.GetForks(ctx, genesisBlockNumber, lastBlockNumber)
if err != nil {
   return []state.ForkIDInterval{}, fmt.Errorf("error getting forks. Please check the configuration. Error: %v", err)
} else if len(forkIntervals) == 0 {
   return []state.ForkIDInterval{}, fmt.Errorf("error: no forkID received. It should receive at least one, please check the configuration...")
}
log.Info("Storing forkID intervals into db")
// Store forkIDs
dbTx, err := st.BeginStateTransaction(ctx)
if err != nil {
   return []state.ForkIDInterval{}, fmt.Errorf("error creating dbTx. Error: %v", err)
}
for _, f := range forkIntervals {
   err := st.AddForkID(ctx, f, dbTx)
   if err != nil {
   err = dbTx.Rollback(ctx)
   if err != nil {
      return []state.ForkIDInterval{}, fmt.Errorf("error when rollback dbTx. Error: %v", err)
   }
      return []state.ForkIDInterval{}, fmt.Errorf("error adding forkID to db. Error: %v", err)
   }
}
err = dbTx.Commit(ctx)
if err != nil {
   return []state.ForkIDInterval{}, fmt.Errorf("error committing dbTx. Error: %v", err)
}
forkIDIntervals = forkIntervals

[agnusmor]

I think is good practice to define always a primary key for the tables, in this case I understand the PK is fork_id

[ARR552]

I don't fully understand what you mean. As you can see the fork_id is defined as Primary key
fork_id BIGINT NOT NULL PRIMARY KEY,