We've got a new update. We now offer our own Web Console. Open the console and type ssh root@your-server-ip to reach any server you own, right from your browser. No SSH app needed.
The Polygon PoS docs list a full node at 8 to 16 cores, 32 to 64 GB of RAM and 4 to 6 TB of storage. That does not fit on a VPS, so here it is a dedicated server from $140 a month. Pay in Bitcoin, Monero, Ethereum or USDT.
Prices are for Romania on a one-month term; a 12-month term takes 50% off.
or $70/mo on a 12-month term (−50%)
Meets the published full-node range (8 to 16 cores, 32 to 64 GB RAM, 4 to 6 TB disk). Stripe the drives into one volume.
Configure and orderor $105/mo on a 12-month term (−50%)
16 TB of raw NVMe for a node you do not want to resize, plus a second copy of a snapshot while you restore.
Configure and orderThe Polygon PoS documentation gives these mainnet figures for a full (sentry) node: minimum 8 cores, 32 GB of RAM and 4 TB of storage; recommended 16 cores, 64 GB and 6 TB, with 1 Gbit/s of network. An archive node needs 16 cores, 64 GB and about 16 TB, and the docs ask for high-IOPS storage for it.
A Polygon PoS node is two programs that run side by side. Heimdall is the consensus layer; Bor is the execution layer (an Ethereum-style client). Both must be installed, both must be in sync, and they talk to each other locally. Current documentation uses Heimdall v2.
The biggest VPS disk here is 1 TB. The Large dedicated server has four 2 TB NVMe drives (8 TB raw) that you stripe into one volume, which covers the recommended 6 TB. The X-Large has four 4 TB drives (16 TB raw) if you want headroom or a second snapshot copy.
Pick the Large dedicated server, or the X-Large if you want headroom.
Join the NVMe drives into one volume mounted at /data.
Use the official install scripts from the Polygon docs, with the versions listed there.
Snapshots cut the sync from weeks to hours. Start Heimdall first, then Bor.
Commands assume Ubuntu 22.04 or 24.04 and a root login. The drive names are examples: run lsblk first. The striping command erases the drives it names.
# log in from your computer
ssh root@YOUR_SERVER_IP
# update the system and create a user that will run the node
apt update && apt -y upgrade
adduser --disabled-password --gecos "" polygon
# firewall: SSH first, then only the ports below
ufw allow OpenSSH
ufw enable
ufw allow 26656/tcp # Heimdall peers
ufw allow 30303/tcp # Bor peers
ufw allow 30303/udp # Bor discovery
# 8545, 8546 and 1317 (RPC/REST) stay closed unless you need themapt -y install mdadm
lsblk
mdadm --create /dev/md0 --level=0 --raid-devices=4 /dev/nvme0n1 /dev/nvme1n1 /dev/nvme2n1 /dev/nvme3n1
mkfs.ext4 /dev/md0
mkdir -p /data && mount /dev/md0 /data
echo "$(blkid -s UUID -o export /dev/md0 | grep ^UUID) /data ext4 defaults,noatime 0 2" >> /etc/fstab
df -h /dataPolygon publishes install scripts and package instructions that change with each release. Copy the current commands, version numbers and the snapshot links from the Polygon full-node docs rather than from an old guide. The overall shape is the same every time:
# 1. install Heimdall (consensus) -> follow the Heimdall v2 steps in the docs
# 2. install Bor (execution) -> follow the Bor steps in the docs
# 3. set the data directories to /data so the big disk is used
# 4. restore the Heimdall and Bor snapshots listed in the docs (hours, not weeks)
# 5. start Heimdall, wait until it is catching up normally, then start Borsystemctl daemon-reload
systemctl enable --now heimdalld
systemctl enable --now bor
journalctl -u heimdalld -f # Ctrl+C to stop watching
journalctl -u bor -fHeimdall exposes a status endpoint on localhost; Bor answers JSON-RPC if you enabled it on localhost.
curl -s localhost:26657/status | grep -o '"catching_up":[a-z]*'curl -s -H 'Content-Type: application/json' \
--data '{"jsonrpc":"2.0","method":"eth_syncing","params":[],"id":1}' http://127.0.0.1:8545
df -h /data # disk left; alert at 85%The Polygon docs list a mainnet full node at 4 TB minimum and 6 TB recommended. An archive node needs about 16 TB. A VPS here tops out at 1 TB, so use the dedicated servers.
Heimdall is the consensus layer and Bor is the execution layer. You run both. They work together on the same server.
26656 TCP for Heimdall peers, and 30303 on both TCP and UDP for Bor peers and discovery. Bor RPC (8545, 8546) and the Heimdall REST API (1317) are optional: open them only if you need to.
Syncing the whole chain from the start takes a very long time. The docs recommend preparing snapshots before you set the node up, which brings the sync down to hours.
No. You create an account with a generated login, top up with crypto and deploy. See no-KYC VPS hosting. You stay responsible for what you run, under our terms.
Checked 20 October 2026. Chain sizes and software requirements change, so confirm against the source before you order.
Pick a server, pay in crypto, log in as root. No ID, no card.
