Apollo Sharding , Did it work?
Mid July, Apollo foundation performed what is believed to be the 1st Database Shard on a active blockchain, so we are here to investigate the claim and take a look at the database, both before, and after the shard.
So what is sharding?
Blockchain operations are optimized to run on code from the last slice of the ledger record. Each slice contains reliable data, and this is ensured by signing its summary table.
This makes downloading the full blockchain unnecessary. You get the latest slice of the ledger — and start processing. Previous slices can be accessed on demand — from last block to first.
The result is remarkable speed and scalability.
So lets take a look….
This is the Apollo database taken before the first shard — Block 2,224,000
Take note of the structure and the size (2.1gb)
This is the database after the first shard, and the wallet upto date at block — Block 2,537,801
Take note of the structure and the size now:
Now below (with the wallet shut-down), I have deleted the first shard! And run up the wallet.
Remembering the Database now only contains blocks from 2,250,000 and above, all previous blocks are now deleted,
And existing database now only 875mb.
Wallet is now in sync, and it has downloaded the few transactions from
the first shard that it needed too, for transactions associated with the current wallet address, but only 7mb worth! Nothing more!
The wallet below has been in sync for another 4hrs,
and only the apl-blockchain.h2.db has changed in size with the increased blocks. INCREDIBLE.
This tells me that we indeed are running on the last slice taken from the ledger, as it does not need all previous blocks to run. Database is just a fraction of the size! Eradicating blockchain bloat and becoming truely sustainable.
