Configuration

Every Snowblossom program is configured by a single Java properties file (key=value, # comments) given as its only command-line argument. The same conventions apply everywhere; the per-program options are documented on the Node, Wallet, Mining and Explorer pages.

Format and environment overrides

Config files are read with duckutil.ConfigFile, a thin wrapper over java.util.Properties. Rules:

• One key=value per line; whitespace around = is tolerated; lines starting with # are comments. Lists are comma-separated (shards=3,4, node_uri=grpc://a,grpc+tls://b).
• Booleans: 1, true, yes, y (case-insensitive) are true; anything else or an unset key is false.
• Environment variables override the file. Every program constructs its config with the prefix snowblossom_, so for any key foo the variable snowblossom_foo wins if it exists. A file may change the prefix with env_override_prefix=…. This is how the Docker images are parameterised and a handy way to run several instances from one file:

export snowblossom_wallet_path=wallets/trading
export snowblossom_network=testnet
java -jar SnowBlossomClient_deploy.jar configs/client.conf balance

• A missing required key aborts with Missing required key: <key>.
• Paths are relative to the working directory, not to the config file.

Selecting the network

The network key picks the parameter set (see Networks & parameters). When it is missing, mainnet is assumed.

Wallet files record their network and refuse to load under a different one; addresses carry a network-specific checksum, so a testnet address cannot be pasted into a mainnet send by accident.

Connecting to a node

Clients, miners, pools and the explorer find their node with one of these (mutually exclusive) settings:

TLS requires a 64-bit JVM; on 32-bit JVMs the plaintext fallback seeds are used. See TLS and node identities.

Logging

Logging uses java.util.logging. Point log_config_file at a properties file such as the shipped configs/logging.properties:

handlers=java.util.logging.ConsoleHandler, java.util.logging.FileHandler
.level=INFO
snowblossom.level=INFO
io.grpc.level=SEVERE
io.netty.level=SEVERE
java.util.logging.ConsoleHandler.level=INFO
java.util.logging.FileHandler.level=FINEST
java.util.logging.FileHandler.pattern = logs/snowblossom-%g-%u.log
java.util.logging.FileHandler.limit = 1000000
java.util.logging.FileHandler.count = 6
java.util.logging.SimpleFormatter.format=[%1$tY-%1$tm-%1$td %1$tH:%1$tM:%1$tS] %4$7s %2$s %5$s%6$s%n

The logs/ directory must exist or the file handler fails at start (“FAILED TO INITIALIZE LOGGING”). Without log_config_file the root level is FINE and gRPC/Netty are silenced. Loggers are named snowblossom.node, snowblossom.peering, snowblossom.userservice, snowblossom.blockchain, snowblossom.db, snowblossom.mempool, snowblossom.client, snowblossom.miner, snowblossom.tls; levels are re-applied to lazily created loggers every 2.5 s, so raising snowblossom.peering.level=FINE works at runtime start.

Example configuration files

The release zip and example/configs/ contain these; they are also the defaults baked into the Docker images.

network=snowblossom
log_config_file=configs/logging.properties
db_type=rocksdb
db_path=node_db/mainnet
#addr_index=true
#tx_index=true
service_port=2338
tls_service_port=2348
tls_key_path=node_db/tls_mainnet
trustnet_key_path=node_db/trustnet

network=snowblossom
log_config_file=configs/logging.properties
node_host=localhost
#node_port=
wallet_path=wallet

network=snowblossom
log_config_file=configs/logging.properties
pool_host=
#pool_port=          # default 23380
snow_path=snow
#auto_snow=true
mine_to_wallet=wallet
#mine_to_address=
#threads=8
#remark=
#memfield=true

log_config_file=configs/logging.properties
node_uri=grpc://localhost,grpc+tls://snow-b.1209k.com
db_type=atomic_file
db_path=pool_db/mainnet
pool_fee=0.005
pool_address=
pay_the_duck=0.01
#remark=
#vote_yes=1,2,3
#vote_no=4
#report_path=report.txt

explorer.conf

node_uri=grpc://localhost:2338
log_config_file=configs/logging.properties
network=snowblossom
port=8080

JVM options

The Bazel targets for SnowBlossomNode and MrPlow pin -Xms4g -Xmx4g; ShackletonExplorer gets 1 GB. When you run a deploy jar yourself you set the heap: java -Xmx4g -jar …. Miners with memfield=true need a heap larger than the field. The Docker images read extra options from SNOWBLOSSOM_JAVA_OPTIONS and add -XX:OnOutOfMemoryError="kill -9 %p" so a container restarts instead of limping on.