Ragno

A simple yet-powerful and updated crawler for Ethereum's EL. It has some basics to crawl and connect to EL nodes.

Installation

# Download the git repository recursively (to update submodules)
git clone --recurse-submodules [email protected]:migalabs/ragno.git && cd ragno

# Install
make dependencies && make install

Usage

The tool supports the following options:

USAGE:
   ragno [commands] [options...]

COMMANDS:
   discv4   discover4 prints nodes in the discovery4 network
   run      run connects to nodes provided in csv file and save into postgresql database
   connect  connect and identify any given ENR
   help, h  Shows a list of commands or help for one command


GLOBAL OPTIONS:
   --help, -h  show help

Environment variables

More specific parameters can be also be configured. When running with Docker, they are set according to your .env file. Otherwise, setting them with flags directly is necessary.

OPTIONS:

--log-level                (string)    Defines the log level of the logs. ("trace", "info", "debug")
--db-endpoint              (string)    Complete endpoint of the database where the recollected data will be saved.
--ip                       (int)       IP to assign to the host.
--port                     (int)       Port to assign to the host.
--metrics-ip               (string)    IP where Prometheus metrics will be hosted.
--metrics-port             (int)       Port where Prometheus metrics will be exposed.
--metrics-endpoint         (string)    Endpoint where Prometheus metrics will be hosted.
--dialers, -cd             (int)       Amount of concurrent dialers for node connections.
--persisters, -cs          (int)       Amount of database writers.
--conn-timeout, -ct        (string)    Time to wait until a connection attempt is considered timed-out.
--snapshot-interval, -si   (string)    How often to insert into the `active_peers` table (snapshots of active nodes).
--ip-api-url, -ipapi       (string)    Full template URL to the API used for retrieving detailed IP information(`ip-api.com`).
--deprecation-time, -dt    (string)    Time limit for reconnecting to nodes before labelling them as deprecated.

Docker

Build and run the database alongside ragno:

These containers are configured with a .env file. See .env.example for examples on the parameters.

docker compose up db ragno

Build and run Prometheus metrics database (exposed at :9090):

This container is configured with a ./prometheus/prometheus.yml file. View ./prometheus/prometheus.yml.example for a template.

docker compose up prometheus

If any container fails to start, make sure you granted permissions to the data folder with: sudo chmod 777 ./app-data/*_db

Prometheus

Prometheus is what is used to gather metrics periodically from the recollected data. By default, they can be viewed at :9070/metrics.

Current available metrics:

name	description
crawler_observed_client_distribution	Number of nodes using the clients seen.
crawler_client_version_distribution	Number of nodes from each of the client's versions.
crawler_geographical_distribution	Number of nodes from each country.
crawler_node_distribution	Total number of nodes.
crawler_deprecated_nodes	Total number of deprecated nodes.
crawler_os_distribution	OS distribution of connected nodes.
crawler_arch_distribution	Architecture distribution of the active nodes in the network.
crawler_hosted_peers_distribution	Distribution of nodes that are hosted on non-residential networks.
crawler_observed_rtt_distribution	Distribution of RTT between the crawler and the nodes in the network.
crawler_observed_ip_distribution	Distribution of IPs hosting nodes in the network.

Migrate

To move between database versions, use go migrate.

In case of any database conflict, you can still force a specific version:

migrate -path / -database "postgresql://username:secretkey@localhost:5432/database_name?sslmode=disable" force <version>

If specific upgrades or downgrades need to be done manually, one could do this with:

migrate -path database/migration/ -database "postgresql://user:password@localhost:5432/database_name?sslmode=disable" -verbose up

Output/Database tables info

node_info

Information about Execution Layer nodes.

column	description
node_id	The node's ID (decoded from the node's record). It is the primary key of the table.
pubkey	The node's secp256k1 public key.
ip	The node's IPv4 address.
tcp	The node's TCP port.
first_connected	Timestamp of when the first successful connection to the node was made.
last_connected	Timestamp of when the last successful connection to the node was made.
last_tried	Timestamp of the last connection attempt.
raw_user_agent	The node's full user agent. (format client/version/os-architechture/language version).
capabilities	The node's capabilities/supported protocols.
error	Latest connection error.
deprecated	Nodes will be marked as deprecated when no connection attempt to it was successful after 48 hours, or if the node is not from mainnet (network ID 1).
client_name	The node's client name.
client_raw_version	The node's full client version (with build info).
client_clean_version	The node's client version.
client_os	Operating system of the node.
client_arch	Computer architecture of the node.
client_language	Language the client of the node is written in.
fork_id	Fork ID the node follows.
protocol_version	Ethereum protocol version the node follows.
head_hash	Hash of the latest block (head) the node sees.
network_id	Network ID of the network where the node resides.
total_difficulty	The total difficulty/cumulative measure of work up to the node.
latency	Time in milliseconds between the latest successful connection attempt and the connection itself.

active_peers

Contains periodical snapshots of active nodes.

column	description
timestamp	Timestamp of when the snapshot was taken.
peers	Array of the nodes active at that moment. The nodes are indexes referencing node_info's id column.

enrs

Contains the response from the Discovery process. This information is used for connection attempts.

column	description
node_id	The node's ID (decoded from the node's record). It is the primary key of the table.
origin	The discovery type the node uses (e.g. discv4).
first_seen	Timestamp of the first time the node was discovered.
last_seen	Timestamp of the last time the node was seen.
ip	The node's IPv4 address.
tcp	The node's TCP port.
udp	The node's UDP port.
seq	The node's record sequence number.
pubkey	The node's secp256k1 public key.
record	The node's record.

ip_info

Contains more detailed information about the node's IP. The data gathered to populate this table is from ip-api.com.

column	description
ip	An IPv4 address. It is the primary key of the table.
expiration_time	Timestamp of the IP's expiration time.
continent	The IP's continent name.
continent_code	The IP's continent ISO 3166-1 alpha-2 code.
country	The IP's country name.
country_code	The IP's country ISO 3166-1 alpha-2 code.
region_name	The IP's region name.
region	The IP's region ISO 3166-2 code.
city	The IP's city name.
zip	The IP's location zip code.
lat	The IP's location's latitude in degrees.
lon	The IP's location's longitude in degrees.
isp	The Internet Service Provider of the IP.
org	The organization that manages the IP address.
as_raw	The full AS (Autonomous System) name the IP is part of.
asname	The AS (Autonomous System) name the IP is part of.
mobile	If the IP is associated with a mobile network or not.
proxy	If the IP is associated with a proxy server.
hosting	If the IP is associated with a hosting provider.

Maintainer

@cortze

Notes

Be careful with the input and output csv file names, since the discv4 command will take output.csv by default.

Also, please note that the tool is currently in a developing stage. Any bugs report and/or suggestions are very welcome.