docs: refresh snapshot operator guidance + tx submission caveats

docs/rpc/transaction/broadcast-tx-commit.mdx

@@ -12,5 +12,6 @@ import FastnearDirectOperation from '@site/src/components/FastnearDirectOperatio
 
 `broadcast_tx_commit` request type
 
+Use a freshly signed base64 transaction for each request. This legacy synchronous send is still documented, but `send_tx` is the preferred path for new integrations, and the sample payload is illustrative only.
 
 <FastnearDirectOperation pageModelId="rpc-broadcast-tx-commit" renderDescription />

docs/rpc/transaction/send-tx.mdx

@@ -12,5 +12,6 @@ import FastnearDirectOperation from '@site/src/components/FastnearDirectOperatio
 
 `send_tx` request type
 
+Use a freshly signed base64 transaction for each request. The sample payload shown in the interactive docs is illustrative only and will fail if you paste it unchanged.
 
 <FastnearDirectOperation pageModelId="rpc-send-tx" renderDescription />

docs/snapshots/index.mdx

@@ -2,7 +2,7 @@
 sidebar_position: 2
 slug: /snapshots
 title: Validator snapshots
-description: Snapshot download paths for FastNear-backed NEAR node bootstrap and recovery workflows.
+description: Current FastNear snapshot status and operator guidance for NEAR node bootstrap and archival snapshot requests.
 sidebar_label: Snapshots Overview
 displayed_sidebar: snapshotsSidebars
 keywords:
@@ -21,53 +21,51 @@ import SimpleButton from '@site/src/components/SimpleButton';
 This section is for node operators who are bootstrapping or recovering NEAR infrastructure. It is not an application-data surface. If the job is reading balances, history, blocks, or contract state, use the API and RPC docs instead of snapshot workflows.
 
 :::warning[Free snapshots are deprecated]
-The free nearcore data snapshots have been deprecated.
+The free nearcore data snapshots were deprecated on June 1, 2025.
 
 The NEAR Infrastructure Committee and Near One recommend Epoch Sync plus decentralized state sync. Use <Link to="https://near-nodes.io">NEAR Nodes</Link> for the current operator guidance and recommended bootstrap posture.
 :::
 
-## Use this section when
+## Current status
 
-- you need to bootstrap a mainnet or testnet node from snapshot data
-- you are recovering an RPC or archival node
-- you already know you want the FastNear snapshot download path
+- FastNear does not currently publish a working public self-serve snapshot download flow from this docs site.
+- For regular RPC nodes and validators, follow <Link to="https://near-nodes.io">NEAR Nodes</Link> and use the current Epoch Sync plus decentralized state sync guidance.
+- For archival snapshot access, visit <Link to="https://fastnear.com/snapshots">fastnear.com/snapshots</Link> or email <Link to="mailto:snapshots@fastnear.com">snapshots@fastnear.com</Link>.
+- The legacy scripts in <Link to="https://github.com/fastnear/static">fastnear/static</Link> still exist, but the old public `latest.txt` discovery flow is no longer the supported public path here.
 
-## Do not use this section when
+## Useful operator references
 
-- you are trying to query chain data for an application
-- you need recent blocks, balances, history, or contract state
-- you are looking for general product API guidance rather than operator setup
+- [NEAR Nodes: Epoch Sync](https://near-nodes.io/intro/node-epoch-sync) for the current boot-node refresh step before first start.
+- [NEAR Nodes: State Sync](https://near-nodes.io/rpc/state-sync) for the current non-snapshot bootstrap recommendation.
+- [NEAR Nodes: Run an Archival Node](https://near-nodes.io/archival/run-archival-node) for archival node configuration and data-directory expectations.
+- [NEAR Nodes: Split Storage for NEAR Archival Nodes](https://near-nodes.io/archival/split-storage-archival) if you are planning hot/cold storage for modern archival setups.
 
-In those cases, use [RPC Reference](/rpc), [FastNear API](/api), [Transactions API](/tx), or [NEAR Data API](/neardata).
-
-## Before you download
+## What to have ready before you request archival access
 
-- Choose the network first: mainnet or testnet.
-- Decide whether you need regular RPC data or archival data.
-- Make sure you understand where hot and cold data must live before starting an archival download.
-- Install `rclone`, because the download scripts depend on it.
+- the network: `mainnet` or `testnet`
+- the node role: validator, RPC node, or archival node
+- whether you expect a single `~/.near/data` restore or a split hot/cold layout
+- the target `NEAR_HOME` path and the storage layout you plan to use
+- any time sensitivity, bandwidth constraints, or recovery deadline that affects the transfer plan
 
-:::info[Getting `rclone`]
-Install `rclone` with:
-
-```bash
-sudo -v ; curl https://rclone.org/install.sh | sudo bash
-```
-:::
-
-## What each path covers
+## Use this section when
 
-- **Mainnet** includes optimized `fast-rpc`, standard RPC, and archival hot/cold download paths.
-- **Testnet** includes RPC and archival snapshot paths for testnet operators.
+- you are confirming the current FastNear snapshot posture for mainnet or testnet
+- you need the right request path for archival snapshot access
+- you are deciding whether to use snapshots or the standard NEAR node bootstrap guidance
 
-See <Link to="https://github.com/near/nearcore?tab=readme-ov-file#about-near">nearcore</Link> for node requirements, and <Link to="https://github.com/fastnear/static">fastnear/static</Link> for the snapshot download script source used by these guides.
+## Do not use this section when
 
-## Need a workflow?
+- you are trying to query chain data for an application
+- you need recent blocks, balances, history, or contract state
+- you are looking for general product API guidance rather than operator setup
 
-Use [Snapshot Examples](/snapshots/examples) for worked operator examples like choosing between optimized `fast-rpc`, standard RPC recovery, and archival hot/cold snapshot paths.
+In those cases, use [RPC Reference](/rpc), [FastNear API](/api), [Transactions API](/tx), or [NEAR Data API](/neardata).
 
 ## Choose a network
 
+Use the network pages below for the current status and escalation path. They no longer assume a public self-serve download URL.
+
 <div style={{display: 'flex', flexWrap: 'wrap', gap: '1rem', margin: '2rem 0'}}>
   <SimpleButton to="/snapshots/mainnet">Mainnet Snapshots</SimpleButton>
   <SimpleButton to="/snapshots/testnet">Testnet Snapshots</SimpleButton>

docs/snapshots/mainnet.mdx

@@ -1,134 +1,46 @@
 ---
 sidebar_label: Mainnet
 slug: /snapshots/mainnet
-description: Download mainnet RPC and archival snapshots for FastNear-backed NEAR infrastructure bootstrapping.
+description: Current FastNear mainnet snapshot posture for RPC bootstrap and archival snapshot requests.
 ---
 
 # Mainnet
 
-## Optimized Mainnet Snapshot
+FastNear no longer documents a public self-serve mainnet snapshot download flow here.
 
-This is likely the preferred approach for syncing, as opposed to downloading an archival snapshot, which is significantly larger and more special-purpose.
-
-Nodes with sufficient resources can take advantage of setting the `$RPC_TYPE` flag to `fast-rpc`. (Default is `rpc`)
-
-Before running the snapshot download script, you can set the following environment variables:
-
-- `CHAIN_ID` to either `mainnet` or `testnet`. (default: `mainnet`)
-- `RPC_TYPE` to either `rpc` (default) or `fast-rpc`
-- `THREADS` to the number of threads you want to use for downloading. Use `128` for 10Gbps, and `16` for 1Gbps (default: `128`).
-- `TPSLIMIT` to the maximum number of HTTP new actions per second. (default: `4096`)
-- `BWLIMIT` to the maximum bandwidth to use for download in case you want to limit it. (default: `10G`)
-- `DATA_PATH` to the path where you want to download the snapshot (default: `~/.near/data`)
-- `BLOCK` to the block height of the snapshot you want to download. If not set, it will download the latest snapshot.
-
-**Run this command to download the RPC Mainnet snapshot:**
-
-:::info
-We will set the following environment variables:
-- `DATA_PATH=~/.near/data` - the standard nearcore path
-- `CHAIN_ID=mainnet` - to explicitly specify the mainnet data
-- `RPC_TYPE=fast-rpc` - select optimized approach
-:::
-
-`RPC Mainnet Snapshot » ~/.near/data`:
-
-```bash
-curl --proto '=https' --tlsv1.2 -sSf https://raw.githubusercontent.com/fastnear/static/refs/heads/main/down_rclone.sh | DATA_PATH=~/.near/data CHAIN_ID=mainnet RPC_TYPE=fast-rpc bash
-```
-
-## RPC Mainnet Snapshot
-
-This is the standard method to obtain a snapshot without the high performance from the previous section covering optimized snapshots.
-
-Before running the snapshot download script, you can set the following environment variables:
-
-- `CHAIN_ID` to either `mainnet` or `testnet`. (default: `mainnet`)
-- `RPC_TYPE` to either `rpc` (default) or `fast-rpc`
-- `THREADS` to the number of threads you want to use for downloading. Use `128` for 10Gbps, and `16` for 1Gbps (default: `128`).
-- `TPSLIMIT` to the maximum number of HTTP new actions per second. (default: `4096`)
-- `BWLIMIT` to the maximum bandwidth to use for download in case you want to limit it. (default: `10G`)
-- `DATA_PATH` to the path where you want to download the snapshot (default: `~/.near/data`)
-- `BLOCK` to the block height of the snapshot you want to download. If not set, it will download the latest snapshot.
-
-**Run this command to download the RPC Mainnet snapshot:**
-
-:::info
-We will set the following environment variables:
-- `DATA_PATH=~/.near/data` - the standard nearcore path
-- `CHAIN_ID=mainnet` - to explicitly specify the mainnet data
+:::warning[Free snapshots are deprecated]
+The free nearcore data snapshots were deprecated on June 1, 2025.
 :::
 
-`RPC Mainnet Snapshot » ~/.near/data`:
-
-```bash
-curl --proto '=https' --tlsv1.2 -sSf https://raw.githubusercontent.com/fastnear/static/refs/heads/main/down_rclone.sh | DATA_PATH=~/.near/data CHAIN_ID=mainnet bash
-```
-
-## Archival Mainnet snapshot
-
-:::warning
-**Time and storage intensive.**
-
-Be prepared for a large download and the inherent time constraints involved.
-
-The snapshot size is ~60Tb and contains more than 1M files.
-:::
+## For regular RPC nodes and validators
 
-Before running the download script, you can set the following environment variables:
+Use [NEAR Nodes](https://near-nodes.io) for the current bootstrap guidance based on Epoch Sync plus decentralized state sync. That is the public recommendation for normal mainnet node bring-up and recovery.
 
-- `CHAIN_ID` to either `mainnet` or `testnet`. (default: `mainnet`)
-- `THREADS` to the number of threads you want to use for downloading. Use `128` for 10Gbps, and `16` for 1Gbps (default: `128`).
-- `TPSLIMIT` to the maximum number of HTTP new actions per second. (default: `4096`)
-- `DATA_TYPE` to either `hot-data` or `cold-data` (default: `cold-data`)
-- `BWLIMIT` to the maximum bandwidth to use for download in case you want to limit it. (default: `10G`)
-- `DATA_PATH` to the path where you want to download the snapshot (default: `/mnt/nvme/data/$DATA_TYPE`)
-- `BLOCK` to the block height of the snapshot you want to download. If not set, it will download the latest snapshot.
+Before the first `neard run`, refresh `network.boot_nodes` with the current command from [NEAR Nodes: Epoch Sync](https://near-nodes.io/intro/node-epoch-sync). A freshly downloaded `config.json` can still leave `boot_nodes` empty, which can leave the node stuck at `Waiting for peers 0 peers`.
 
-By default, the script assumes the paths for the data:
-- Hot data (has to be on NVME): `/mnt/nvme/data/hot-data`
-- Cold data (can be on HDDs): `/mnt/nvme/data/cold-data`
+## For archival mainnet snapshots
 
+Archival snapshot access is request-based. Start from [fastnear.com/snapshots](https://fastnear.com/snapshots) or email [snapshots@fastnear.com](mailto:snapshots@fastnear.com) to get the current availability, block height, and storage expectations.
 
-**Run the following commands to download the Archival Mainnet snapshot:**
+## About the legacy scripts
 
-1. Download the latest snapshot block height:
+The [fastnear/static](https://github.com/fastnear/static) repository still contains the historical `down_rclone.sh` and `down_rclone_archival.sh` scripts. These docs no longer publish copy-paste commands for them because the old public `latest.txt` discovery URLs now redirect to the snapshots info page and are not a supported public workflow.
 
-`Latest archival mainnet snapshot block`:
+If FastNear support gives you a current block height and a supported download path, use the exact host, block, and storage layout they provide instead of the retired public examples.
 
-```bash
-LATEST=$(curl -s "https://snapshot.neardata.xyz/mainnet/archival/latest.txt")
-echo "Latest snapshot block: $LATEST"
-```
+## If you are preparing for an archival request
 
-2. Download the HOT data from the snapshot. It has to be placed on NVME.
+- Be ready to separate hot and cold data if the snapshot package requires it.
+- Confirm your storage layout with nearcore before you start the transfer.
+- Install `rclone` only if the support flow you receive actually depends on it.
 
-:::info
-We will set the following environment variables:
-- `DATA_TYPE=hot-data` - downloads the Hot data
-- `DATA_PATH=~/.near/data` - the standard nearcore path
-- `CHAIN_ID=mainnet` - to explicitly specify the mainnet data
-- `BLOCK=$LATEST` - specify the snapshot block
-:::
-
-`Archival Mainnet Snapshot (hot-data) » ~/.near/data`:
-
-```bash
-curl --proto '=https' --tlsv1.2 -sSf https://raw.githubusercontent.com/fastnear/static/refs/heads/main/down_rclone_archival.sh | DATA_TYPE=hot-data DATA_PATH=~/.near/data CHAIN_ID=mainnet BLOCK=$LATEST bash
-```
-
-3. Download the COLD data from the snapshot. It can be placed on HDDs.
-
-:::info
-We will set the following environment variables:
-- `DATA_TYPE=cold-data` - downloads the Hot data
-- `DATA_PATH=/mnt/hdds/cold-data` - the path where to place cold data. **Note**: the nearcore config should point cold data store to the same path.
-- `CHAIN_ID=mainnet` - to explicitly specify the mainnet data
-- `BLOCK=$LATEST` - specify the snapshot block
-:::
+## If FastNear provides a snapshot package
 
-`Archival Mainnet Snapshot (cold-data) » /mnt/hdds/cold-data`:
+1. Stop `neard` before you replace any snapshot data or edit archival storage settings.
+2. Confirm whether the package is meant for `~/.near/data` or for split `hot-data` and `cold-data` paths.
+3. Verify that your `config.json` matches the intended archival mode before restart.
+   For archival nodes, NEAR Nodes documents `archive: true` and `tracked_shards: [0]` as critical settings.
+4. If the package uses split storage, make sure your configured store paths match the delivered hot/cold layout before you start the node again.
+5. Restart the node and verify that it starts cleanly and resumes syncing instead of silently serving stale local state.
 
-```bash
-curl --proto '=https' --tlsv1.2 -sSf https://raw.githubusercontent.com/fastnear/static/refs/heads/main/down_rclone_archival.sh | DATA_TYPE=cold-data DATA_PATH=/mnt/hdds/cold-data CHAIN_ID=mainnet BLOCK=$LATEST bash
-```
+For nearcore storage expectations and operator prerequisites, see [nearcore](https://github.com/near/nearcore?tab=readme-ov-file#about-near).