Golem SDK

Provider CLI reference

Golem Provider Command Line Interface
Keep in mind that that is a really basic release, and future updates will give you much more options to play with.

CLI commands

Run golemsp help without arguments to see top-level usage information:
$ golemsp help
golemsp 0.12.3 (9de9778d 2023-08-01 build #309)
User friedly CLI for running Golem Provider
golemsp <SUBCOMMAND>
-h, --help Prints help information
-V, --version Prints version information
run Run the golem provider
stop Stop the golem provider
settings Manage settings
status Show provider status
help Prints this message or the help of the given subcommand(s)


golemsp run - Start the Golem provider. It uses mainnet as a payment platform by default.
Invoke golemsp run --help to see more options.


$ golemsp settings
golemsp-settings 0.3.0
Manage settings
golemsp settings <SUBCOMMAND>
-h, --help Prints help information
-V, --version Prints version information
set Change settings
show Show current settings
help Prints this message or the help of the given subcommand(s)

Settings set

golemsp settings set --help - to see how to change settings. Invoking this will prompt usage, flags, and options.
$ golemsp settings set --help
golemsp-settings-set 0.3.0
Change settings
golemsp settings set [OPTIONS]
-h, --help Prints help information
-V, --version Prints version information
--node-name <node-name>
--cores <num> Number of shared CPU cores
--memory <bytes (like "1.5GiB")> Size of shared RAM
--disk <bytes (like "1.5GiB")> Size of shared disk space
--starting-fee <GLM (float)> Price for starting a task
--env-per-hour <GLM (float)> Price for working environment per hour
--cpu-per-hour <GLM (float)> Price for CPU per hour
--account <account> Account for payments [env: YA_ACCOUNT=]
--payment-network <network> Payment network [env: YA_PAYMENT_NETWORK_GROUP=] [default: mainnet]
[possible values: mainnet, testnet]
In order to change a particular setting (for eg. price settings) type:
golemsp settings set --cpu-per-hour 3
You can also combine multiple settings in one command as follows:
golemsp settings set --cpu-per-hour 3 --cores 7 which will change your GLM per hour to "3" and adjust the numbers of shared CPU cores to "7".
To change the default Ethereum address that was created for you during the initial setup process type:
golemsp settings set --account <address>
and restart your node afterwards for it to update. To check if your address has been updated properly run golemsp status

Settings show

golemsp settings show - Show current settings.
$ golemsp settings show
node name: "zima"
Shared resources:
cores: 7
memory: 10.604358091950417 GiB
disk: 59.39304809570313 GiB
Pricing for preset "vm":
0.025000000000000001 GLM per cpu hour
0.005000000000000000 GLM per hour
0.000000000000000000 GLM for start
Pricing for preset "wasmtime":
0.025000000000000001 GLM per cpu hour
0.005000000000000000 GLM per hour
0.000000000000000000 GLM for start


golemsp status - Print out the status of your node.
When the node is not running you'll see:
$ golemsp status
│ Status │
│ │
│ Service is not running │
│ Version 0.12.3 │
│ Commit 9de9778d │
│ Date 2023-08-01 │
│ Build 309 │
│ │
│ Node Name lato │
│ Subnet public │
│ VM valid │
When your node is already running golemsp status will show:
$ golemsp status
│ Status │ Wallet │ Tasks │
│ │ 0x2a14f8ae0272bd4c38ed1b40c66e88ed719dab69 │ │
│ Service is running │ │ last 1h processed 0 │
│ Version 0.12.3 │ network mainnet │ last 1h in progress 0 │
│ Commit 9de9778d │ amount (total) 0 GLM │ total processed 509 │
│ Date 2023-08-01 │ (on-chain) 0 GLM │ (including failures) │
│ Build 309 │ (polygon) 0 GLM │ │
│ │ (zksync) 0 GLM │ │
│ Node Name lato │ │ │
│ Subnet public │ pending 0 GLM (0) │ │
│ VM valid │ issued 0 GLM (0) │ │
In the three columns, you can check the basic information regarding the status of your node


  • Whether your node is running
  • Version of your node (with commit, build date and build number)
  • Name of your node
  • Subnet in which your node is currently running
  • VM status


  • Account address
  • Payment network: mainnet or rinkeby
  • Amount of tokens that you have earned for successful computation
  • On-chain amount of tokens that you have earned (explorer etherscan.io or rinkeby.etherscan.io)
  • Zk-sync amount of tokens that you have earned (explorer zkscan.io or rinkeby.zkscan.io)
  • Pending payments that you should receive for computation
  • Amount of tokens that is still unconfirmed and may not show on your account


  • Number of tasks that you were computing in last hour
  • Number of tasks that were in progress during the last hour
  • Total task that you were trying to compute - including those that were not computed

Exit GLM tokens to Ethereum

While not specific to the provider CLI, at some point you may want to move your tokens. By default, mainnet tasks are paid on Layer 2. Assuming you have a local wallet, you can interact with the payment driver to exit your tokens from Layer 2 to Layer 1. This is done using theyagna payment exit command. With this command there are two main flags to keep in mind; --networkand --to-address.
For --networkyou have two options, either mainnet or rinkeby. For --to-addressyou can specify a destination address other than the local wallet address.
To exit your GLM to the same address on Ethereum mainnet type:
yagna payment exit --network=mainnet
To exit your GLM to a different address on Ethereum mainnet type:
yagna payment exit --to-address=<address> --network=mainnet
Note that if you decided to use an external wallet during your setup process, you can connect to ZkSync's wallet at https://wallet.zksync.io/ and exit that way. In the scenario that a different payment driver is being used, you will need to use the relevant available options to connect and access your tokens.

Advanced Settings

ya-provider allows to fine-tune the config created with golemsp settings using commands config, preset, profile, and exe-unit.
Additionally, it enables configuration of the certificate keystore and of the domain whitelist, which are used to control what kind of outbound traffic is allowed out of the VM containers run by the requestors on your machine.


The provider has an embedded certificate keystore which is used to validate any additional permissions for the payload launched by the requestors.
By default it contains only Golem public certificate which allows to execute example app and apps from trusted by Golem creators (certificates allow to verify incoming Demand's Computational Payload Manifests).
Run ya-provider keystore --help to see possible subcommands

Domain whitelist

The Computational Payload Manifests embedded in the Demands can specify a list of URLs which may get called by the services running on a Provider. If the manifest declares requests to URLs in domains that are not whitelisted, it must come with a signature and app author's public certificate. By default, the domains whitelist consists of a curated set of public websites and APIs like github, dockerhub or public Ethereum nodes.
Run ya-provider whitelist --help to see possible subcommands