GPU assisted Proof of Capacity (PoC) Miner for Burstcoin (BURST)
- edit 'jminer.properties' with text editor to configure miner
- ensure java8 (64bit) and openCL driver/sdk is installed (java9 will not work)
- execute 'java -jar -d64 -XX:+UseG1GC burstcoin-jminer-0.6.x-RELEASE.jar' or run the .bat/.sh file
'-d64' to ensure 64bit java (remove for 32bit) '-XX:+UseG1GC' to free memory after round finished.
The miner is configured in a text-file named 'jminer.properties'. This file has to be in the miner directory (same folder as '*.jar' file)
To get started, use one of the following examples. The min. required settings for the different mining-modes.
plotPaths=D:/,C:/,E:/plots,F:/plots
numericAccountId=<YOUR NUMERIC ACCOUNT ID>
poolServer=http://pool.com:port
plotPaths=D:/,C:/,E:/plots,F:/plots
poolMining=false
passPhrase=<YOUR PASS PHRASE>
jminer still supports POC1
only use one type, POC1 or POC2 on one drive ('plotPath'), mixed will be skipped for now.
ensure your POC2 plotfiles do not have staggersize in filename, or they will be treated like POC1.
POC1 filename: 'numericAccountId_startNonce_numberOfNonces_staggersize'
POC2 filename: 'numericAccountId_startNonce_numberOfNonces'
your 'jminer.properties' hasn't got to contain all properties listed here, most of them are optional or there is a fallback/default value for it.
list of plot paths separated with , e.g. D:/,C:/,E:/plots,F:/plots (in one line) the miner will treat every path as 'physical' drive and use one thread for it
plotPaths=D:/,C:/,E:/plots,F:/plots
optional 'true' will check 'plotPaths' for changed plot files on every round 'false' will check only on start/restart if you are moving/creating plot-files while mining, it could be disabled
scanPathsEveryRound=false
optional ... list all plotFiles on start. If walletServer/soloServer is configured, it will show mined blocks and drive seeks/chunks of plotfile, too.
listPlotFiles=true
'true' for pool mining, 'false' for solo mining. ensure to configure the chosen mining-mode below. For solo-mining you need to set
poolMining=false
min. deadline to be committed. Will be used for pool mining if 'forceLocalTargetDeadline=true'
targetDeadline=750000
'true' will force jminer to use the targetDeadline specified below, even if the pool says otherwise. Only for pool mining! 'false' uses the default behavior for targetDeadline. https://www.youtube.com/watch?v=9lwogE31SiI
forceLocalTargetDeadline=true
'true' will overrule 'targetDeadline' and 'forceLocalTargetDeadline' settings. the miner will calculate the targetDeadline dynamic on poolMining.
dynamicTargetDeadline=true
Ensure you already setup reward assignment http://localhost:8125/rewardassignment.html
first number in all plot-files
numericAccountId=xxxxxxxxxxxxxxx
format is inclusive protocol and port e.g. 'http://pool.com:8125'
poolServer=http://pool.com
define local or online wallet, to receive and show last winner! if empty, winner feature will be just disabled. format is inclusive protocol and port e.g. 'http://localhost:8125' online use e.g. 'https://wallet.burst-team.us:8125'
walletServer=https://wallet.burst-team.us:8125
number of retries to get winner from walletServer
winnerRetriesOnAsync=10
time to wait until next retry to get winner from walletServer
winnerRetryIntervalInMs=250
soloServer (default:http://localhost:8125)
WARN! soloServer should be http://localhost:8125 or http://127.0.0.1:8125 Solo means you send your PASS on commit results!
DO NOT try to use a online wallet or pool as Server!
soloServer=http://127.0.0.1:8125
secretPhrase/password of solo mining burst-account
passPhrase=xxxxxxxxxxxxxx
The miner uses openCL for most of the mining calculations, ensure it is setup correctly. Instructions can be found e.g. here (thanks cryo): https://github.com/bhamon/gpuPlotGenerator/blob/master/README.md You could also use that instruction to find your platformId and deviceId if needed. Since version 0.4.4 all available platforms and devices are listed on startup.
Since version 0.6.0 openCL is optional. Thanks harry1453! Disable (useOpenCl=false) to only use CPU for mining. Following settings are only relevant if openCL is used.
default=false
id of openCL platform on your system. one platform may have multiple devices, the miner currently uses just one (in general not the bottleneck)
platformId=0
specifies the device used by OCLCecker, can be your first GPU, in most cases it will not be 100% used. (depends on capacity)
deviceId=1
interval of asking wallet/pool for mining info (in ms), to check for new block
refreshInterval=2000
restart round on new generationSignature for same round as long as equal miningInfo was not already finished (it may happen that miningInfo changes, and changes back later) false value will disable this feature
updateMiningInfo=false
increase the 'connectionTimeout' on network problems. this timeout is used for all network requests. if you use pool or online-wallet, the 12 sec. default may cause timeout on committing nonces or getting mining info etc.
connectionTimeout=12000
setting 'debug' to true will log additional information of the mining process, that are not related to mining, but to miner internals.
debug=true
setting 'writeLogFile' to 'true' will write all logs from console to a file, too. the name of that file can be specified by 'logFilePath'.
writeLogFile=true
path (filename and optional directory, relative to miner location)
logFilePath=mylogs/jminier/log.txt
defines how often the mining progress is shown per round thats the 'xx% done ...' info.
readProgressPerRound=16
switch between decimal units (true): TB/GB/MB (divided by 1000), or binary units (false) TiB/GiB/MiB (divided by 1024) - https://en.wikipedia.org/wiki/Byte
byteUnitDecimal=false
set this to 'true' to show info about every drive on finish reading it, this is useful to find the slow ones ... can help to optimize your setup.
showDriveInfo=true
e.g. you see in logs, that 'drive-c' and 'drive-d' slow down this mining setup:
read 'C:/data/drive-a' (3TB 958GB) in '28s 444ms'
read 'C:/data/drive-b' (3TB 932GB) in '29s 114ms'
read 'C:/data/drive-c' (3TB 996GB) in '35s 390ms'
read 'C:/data/drive-d' (3TB 996GB) in '35s 685ms'
patterns for logfile and cosole output can be different only needed if you want to configure your own log pattern e.g. following would show colored time and message on console:
logPatternConsole=%blue(%d{HH:mm:ss.SSS}) %green(%msg%n)
only time and message for logfile:
logPatternFile=%d{HH:mm:ss.SSS} %msg%n
For all options please read from docs: https://logback.qos.ch/manual/layouts.html#ClassicPatternLayout
set this to 'true' to show found deadlines below targetDeadline configure or provided by pool (overwriting the targetDeadline specified in jminer.properties)
showSkippedDeadlines=false
staggerSize defines number of nonces per chunk. the miner will split chunks in smaller pieces called chunkParts. this makes sense, to save memory and optimize speed. in the best case chunkPart#1 will be checked before chunkPart#2 is completely read ... depending on the power of your GPU. if staggersize is smaller than chunkPartNonces, staggersize will be used. e.g. play with +/- 160000 steps
chunkPartNonces=960000
normally '0' means, the miner takes one thread per drive (plotPath) this is recommend. choosing a other number of 'readerThreads' can be useful on memory issues. For example, if you mine on 4 drives (plotPaths), you can reduce the memory usage by setting 'readerThreads=2', this will reduce mining speed but save memory.
readerThreads=10