bscwatch
Updated 15:14 UTC

BNB Smart Chain network monitor

Nodes, clients, hosting, validators and block builders, measured from the peer-to-peer network. Last 24 hours, refreshed every 15 minutes.

Nodes seen
2,244
BSC mainnet, key-rotating fleets excluded
Countries
46
with at least one node
Nodes on three hosting providers
48%
Hetzner Online, Amazon.com, OVH
Synced
91%
within 50 blocks of the chain head

Where the nodes are

Nodes per city; bubble area follows the number of nodes. Zoom with the buttons, a double-click or ⌘/Ctrl + scroll, then drag to move. IP Geolocation by DB-IP.

Countries

United States598 · 26.6%
Germany488 · 21.7%
Japan246 · 11.0%
Lithuania158 · 7.0%
Finland108 · 4.8%
Singapore104 · 4.6%
Ireland85 · 3.8%
China65 · 2.9%
Netherlands50 · 2.2%
France49 · 2.2%
United Kingdom42 · 1.9%
Hong Kong42 · 1.9%
Canada36 · 1.6%
Russia26 · 1.2%
South Korea22 · 1.0%
Other125 · 5.6%

Hosting providers

Hetzner Online GmbH426 · 19.0%
Amazon.com, Inc.410 · 18.3%
OVH SAS246 · 11.0%
UAB "Atlantis Capital"158 · 7.0%
TeraSwitch Networks Inc.89 · 4.0%
Latitude.sh55 · 2.5%
UAB Cherry Servers52 · 2.3%
Limestone Networks, Inc.44 · 2.0%
The Constant Company, LLC35 · 1.6%
Interserver, Inc33 · 1.5%
Allnodes Inc30 · 1.3%
Alibaba (US) Technology Co., Ltd.29 · 1.3%
Google LLC28 · 1.2%
Chinanet19 · 0.8%
ReliableSite.Net LLC18 · 0.8%
Other572 · 25.5%

Clients and sync

Clients

geth1,698 · 75.7%
reth-bsc321 · 14.3%
No client name155 · 6.9%
meganode24 · 1.1%
reth16 · 0.7%
gnode10 · 0.4%
Other20 · 0.9%

Client versions

geth v1.7.8768 · 37.2%
geth v1.7.7362 · 17.6%
geth v1.4.13190 · 9.2%
geth v1.8.1172 · 8.3%
reth-bsc v0.1.3147 · 7.1%
geth v1.8.0124 · 6.0%
reth-bsc v0.1.273 · 3.5%
reth-bsc v0.1.155 · 2.7%
geth v1.7.340 · 1.9%
meganode v1.7.724 · 1.2%
Other107 · 5.2%

Sync state

Synced (≤ 50 blocks)2,049 · 91.3%
Behind 50–100k33 · 1.5%
Behind > 100k104 · 4.6%
Unknown58 · 2.6%

Fork

Pasteur2,122 · 94.7%
Maxwell40 · 1.8%
Osaka+Mendel33 · 1.5%
Genesis17 · 0.8%
Fermi13 · 0.6%
Other15 · 0.7%

Open JSON-RPC: 50 nodes over HTTP, 38 over WebSocket. Bootnodes up: 6 of 6.

Unusual nodes

Key-rotating fleet

15 IP addresses showed 36,531 different node identities in 24 hours, a new one every few seconds. They are left out of every other count on this page.

OVH SAS15 · 100.0%

Old version, current fork

190 synced nodes on the current fork announce a client version too old to follow it, on 2 networks. The announced version is not what they run.

UAB "Atlantis Capital"158 · 83.2%
Amazon.com, Inc.32 · 16.8%

Validator nodes

37 validators registered 74 node identities on-chain for the Enhanced Validator Network. 3 of them could be found through public discovery in the last 24 hours; the rest stay hidden behind sentry nodes.

Validators and block builders

Validators take turns producing blocks; almost all blocks are assembled by an external builder. The builder is identified by its payment transaction at the start of the block (48Club: a transfer from 0x4848…4848; BlockRazor: a payment to its builder account 0xc7d4…db5c).

48ClubBlockRazorOther
48Club 57%BlockRazor 42%
12,730 blocks with a full copy received by our observers.

Active validators

Each of the 21 validators in the active set produces about the same number of blocks; stake decides who is in the set, not how often it produces.

Active validators: blocks, builders and transactions public before the block
ValidatorBlocksBuilt byPublic before block
BNBEve4.4%BlockRazor 57% · 48Club 41%29%
LegendIII4.4%48Club 55% · BlockRazor 44%26%
Legend4.4%48Club 55% · BlockRazor 44%25%
NodeReal4.4%48Club 56% · BlockRazor 44%25%
Figment4.3%48Club 62% · BlockRazor 38%27%
LegendII4.3%48Club 54% · BlockRazor 45%25%
Fuji4.3%48Club 60% · BlockRazor 39%25%
Avengers4.2%48Club 62% · BlockRazor 38%27%
Namelix4.2%48Club 55% · BlockRazor 45%27%
TWStaking4.2%48Club 57% · BlockRazor 43%25%
InfStones4.2%48Club 59% · BlockRazor 41%25%
Diablo4.2%48Club 58% · BlockRazor 42%25%
HashKey4.1%48Club 56% · BlockRazor 42%28%
BscScan4.1%48Club 59% · BlockRazor 41%26%
CertiK4.1%48Club 59% · BlockRazor 40%27%
MathW4.1%48Club 52% · BlockRazor 48%26%
Zen4.0%48Club 51% · BlockRazor 49%27%
Shannon4.0%48Club 65% · BlockRazor 35%23%
Feynman4.0%48Club 66% · BlockRazor 34%24%
Defibit3.9%48Club 51% · BlockRazor 48%25%
The48Club3.8%48Club 61% · BlockRazor 39%25%
Candidate validators (24), 12% of blocks
Candidate validators
ValidatorBlocksBuilt byPublic before block
ListaDAO0.7%48Club 58% · BlockRazor 38%27%
Coda0.7%48Club 53% · BlockRazor 47%26%
NovaX0.6%48Club 60% · BlockRazor 40%26%
Legend80.6%48Club 59% · BlockRazor 41%22%
Pexmons0.6%48Club 55% · BlockRazor 45%27%
Veri0.6%48Club 69% · BlockRazor 31%24%
Star0.6%––
Axion0.5%48Club 55% · BlockRazor 45%29%
Ciscox0.5%48Club 58% · BlockRazor 42%24%
LegendVII0.5%48Club 55% · BlockRazor 45%29%
ListaDAO20.5%48Club 55% · BlockRazor 43%23%
Sigm80.5%48Club 61% · BlockRazor 38%23%
Aoraki0.5%48Club 62% · BlockRazor 38%18%
Nexa0.5%48Club 57% · BlockRazor 43%27%
Ankr0.5%48Club 74% · BlockRazor 26%29%
Turing0.5%––
LegendV0.5%48Club 64% · BlockRazor 36%25%
Kraken0.5%––
Seoraksan0.4%48Club 51% · BlockRazor 49%30%
GlobalStk0.4%48Club 58% · BlockRazor 42%25%
Coinlix0.4%48Club 61% · BlockRazor 39%34%
LegendIV0.4%48Club 62% · BlockRazor 38%25%
ListaDAO30.4%48Club 63% · BlockRazor 29%23%
BitLV0.3%48Club 73% · BlockRazor 23%23%

Private order flow

26%

of the transactions in BSC blocks were public before their block.

The rest first appeared in the peer-to-peer network when the block was sealed: they reached the block builder privately. Measured over 9,748 blocks and 796,883 transactions. Plain transfers, approvals and builder payments are left out. A lower bound: each observer holds about 30 peer connections.

How this is measured

  • A crawler walks the BSC discovery network and completes a handshake with every node it can reach. Two observers in different regions keep short, rotating sessions with about 30 peers each and record when every block and transaction first arrives.
  • “Public before block” counts a transaction when an observer received it from the network before the block's timestamp (BSC keeps the milliseconds in the header's mixDigest). Observers see part of the network, so the true share is somewhat higher.

More detail

For more information or more detailed statistics, write to qwart279@gmail.com.