Crate data_anchor

Data Anchor CLI

This crate provides a command line interface for interacting with the blober program and the Data Anchor indexer. It can create and manage blober accounts, upload blobs and query either the ledger or the indexer for uploaded data.

# Basic invocation
data-anchor [GLOBAL-OPTIONS] <SUBCOMMAND> [COMMAND OPTIONS]

The examples below assume the binary is called data-anchor but the application name will match the crate name when built locally.

Global options

Some options are required for almost every command. They may also be set via environment variables. Unless stated otherwise the arguments accept string values.

Flag/Env var	Description
-p, --program-id / DATA_ANCHOR_PROGRAM_ID	Address of the deployed blober program (required)
-n, --namespace / DATA_ANCHOR_NAMESPACE	Namespace used to derive the blober PDA
-b, --blober-pda / DATA_ANCHOR_BLOBER_PDA	Explicit blober PDA to use instead of deriving from the namespace
-s, --payer / DATA_ANCHOR_PAYER	Path to the keypair used to sign transactions
-o, --output / DATA_ANCHOR_OUTPUT	Output format: text, json, json-pretty, csv (default: text)
-i, --indexer-url / DATA_ANCHOR_INDEXER_URL	URL of the indexer service for read operations
--indexer-api-token / DATA_ANCHOR_INDEXER_API_TOKEN	Optional API token for the indexer
-c, --config-file / DATA_ANCHOR_SOLANA_CONFIG_FILE	Solana CLI config file (default: ~/.config/solana/cli/config.yml)

--program-id and either --namespace or --blober-pda must be supplied. The payer defaults to the keypair configured in the Solana config file.

Subcommands

blober (br)

Manage a blober PDA.

data-anchor -p <program> -n <namespace> blober initialize

Example commands:

# Create the PDA for a new namespace
data-anchor -p <PROGRAM_ID> -n <my-namespace> blober initialize

# Close the PDA and reclaim rent
data-anchor -p <PROGRAM_ID> -n <my-namespace> blober close

Commands:

• initialize (i) – create the PDA for the namespace.
• close (c) – close the PDA and reclaim rent.

blob (b)

Upload, discard or retrieve blobs.

Commands:

• upload (u) – upload data. Use -d, --data-path <PATH> to read from a file or --data <HEX> for inline hex data. Without either, data is read from stdin.
• discard (d) – discard a blob using its PDA: data-anchor b d <BLOB_PUBKEY>.
• fetch (f) – fetch blob data from the ledger using transaction signatures: data-anchor b f <SIG> [SIG ...].
• get (g) – retrieve all blobs finalized at a slot. Requires the slot number and optionally -l, --lookback-slots <SLOTS> to limit how far back to search.

Example commands:

# Upload data from a file
data-anchor -p <PROGRAM_ID> -n <my-namespace> blob upload -d ./data.txt

# Discard an unfinished upload
data-anchor -p <PROGRAM_ID> -n <my-namespace> blob discard -b <BLOB_PUBKEY>

# Fetch blob data from transaction signatures
data-anchor -p <PROGRAM_ID> -n <my-namespace> blob fetch <SIG1> <SIG2>

# Get all blobs finalized at a slot using a lookback window
data-anchor -p <PROGRAM_ID> -n <my-namespace> blob get 54321 -l 10

indexer (i)

Query the indexer for blobs or proofs.

Commands:

• blobs (b) <slot> – return blobs finalized at the given slot.
• blobs-for-blober (bl) – requires --blober <PUBKEY> and optional --start <RFC3339>/--end <RFC3339> to restrict the time range.
• blobs-for-payer (bp) – requires -y, --blob-payer <PUBKEY> and -m, --network-name <NAME> with optional time range (--start/--end).
• blobs-for-network (bn) – requires -m, --network-name <NAME> and optional time range.
• blobs-for-namespace (ns) – requires --namespace <NAME> and optional --payer-pubkey <PUBKEY> plus the time range arguments.
• proof-for-blob (pb) – --blob <PUBKEY> returns the proof for the blob.
• proof (p or proofs) <slot> – get the compound proof for a slot.

Example commands:

# Get blobs for a slot via the indexer
data-anchor -p <PROGRAM_ID> -n <my-namespace> indexer blobs 12345

# Query blobs for a specific blober PDA in a time range
data-anchor indexer blobs-for-blober --blober <BLOBER_PUBKEY> \
    --start 2025-06-01T00:00:00Z --end 2025-06-30T00:00:00Z

# Query blobs paid for by an address on a network
data-anchor indexer blobs-for-payer -y <PAYER_PUBKEY> -m <NETWORK_NAME> \
    --start 2025-06-01T00:00:00Z --end 2025-06-30T00:00:00Z

# Query all blobs uploaded on a network in a time range
data-anchor indexer blobs-for-network -m <NETWORK_NAME> \
    --start 2025-06-01T00:00:00Z --end 2025-06-30T00:00:00Z

# Get blobs for a namespace, optionally filtering by payer
data-anchor indexer blobs-for-namespace --namespace <NAME> \
    --payer-pubkey <PAYER_PUBKEY> --start 2025-06-01T00:00:00Z \
    --end 2025-06-30T00:00:00Z

# Fetch the proof for a blob
data-anchor indexer proof-for-blob --blob <BLOB_PUBKEY>

# Get the compound proof for a slot
data-anchor -p <PROGRAM_ID> -n <my-namespace> indexer proof 12345

Time range arguments expect RFC3339 timestamps (e.g. 2025-06-01T00:00:00Z).

benchmark (m)

Tools for generating data and measuring performance.

Commands:

• generate (g) <DATA_PATH> – create random data files. Options:
  • -s, --size <BYTES> size of each file (default: 1000).
  • -c, --count <N> number of files to generate (default: 100).
  • -r, --random-length generate files with random lengths up to size.
• measure (m) <DATA_PATH> – upload all files in the directory and measure throughput. Options:
  • -t, --timeout <SECS> upload timeout per blob (default: 60).
  • -c, --concurrency <N> number of concurrent uploads (default: 600).
  • -p, --priority <LEVEL> transaction priority (min, low, medium, high, very-high).
• automate (a) – run a set of benchmark scenarios.
  • -d, --data-path <PATH> directory used for generated data.
  • -r, --running-csv <FILE> write intermediate results to this CSV file.

Example commands:

# Generate 100 files of 1KiB
data-anchor benchmark generate ./bench-data -s 1024 -c 100

# Measure upload throughput with 600 concurrent uploads
data-anchor benchmark measure ./bench-data -c 600 -t 60 -p medium

# Run a full benchmark scenario
data-anchor benchmark automate -d ./bench-data -r results.csv

The help command or --help flag on any subcommand shows these options at runtime.

Structs

Options