cosmos-exporter

cosmos-exporter is a Prometheus scraper that fetches the data from a full node of a Cosmos-based blockchain via gRPC.

What can I use it for?

You can run a full node, run cosmos-exporter on the same host, set up Prometheus to scrape the data from it (see below for instructions), then set up Grafana to visualize the data coming from the exporter and probably add some alerting. Here are some examples of Grafana dashboards we created for ourselves:

How can I set it up?

First of all, you need to download the latest release from the releases page. After that, you should unzip it and you are ready to go:

wget <the link from the releases page>
tar xvfz cosmos-exporter-*
./cosmos-exporter

That's not really interesting, what you probably want to do is to have it running in the background. For that, first of all, we have to copy the file to the system apps folder:

sudo cp ./cosmos-exporter /usr/bin

Then we need to create a systemd service for our app:

sudo nano /etc/systemd/system/cosmos-exporter.service

You can use this template (change the user to whatever user you want this to be executed from. It's advised to create a separate user for that instead of running it from root):

[Unit]
Description=Cosmos Exporter
After=network-online.target

[Service]
User=<username>
TimeoutStartSec=0
CPUWeight=95
IOWeight=95
ExecStart=cosmos-exporter
Restart=always
RestartSec=2
LimitNOFILE=800000
KillSignal=SIGTERM

[Install]
WantedBy=multi-user.target

Then we'll add this service to the autostart and run it:

sudo systemctl enable cosmos-exporter
sudo systemctl start cosmos-exporter
sudo systemctl status cosmos-exporter # validate it's running

If you need to, you can also see the logs of the process:

sudo journalctl -u cosmos-exporter -f --output cat

How can I scrape data from it?

Here's the example of the Prometheus config you can use for scraping data:

scrape-configs:
  # specific validator(s)
  - job_name:       'validator'
    scrape_interval: 15s
    metrics_path: /metrics/validator
    static_configs:
      - targets:
        - <list of validators you want to monitor>
    relabel_configs:
      - source_labels: [__address__]
        target_label: __param_address
      - source_labels: [__param_address]
        target_label: instance
      - target_label: __address__
        replacement: <node hostname or IP>:9300
  # specific wallet(s)
  - job_name:       'wallet'
    scrape_interval: 15s
    metrics_path: /metrics/wallet
    static_configs:
      - targets:
        - <list of wallets>
    relabel_configs:
      - source_labels: [__address__]
        target_label: __param_address
      - source_labels: [__param_address]
        target_label: instance
      - target_label: __address__
        replacement: <node hostname or IP>:9300

  # all validators
  - job_name:       'validators'
    scrape_interval: 15s
    metrics_path: /metrics/validators
    static_configs:
      - targets:
        - <node hostname or IP>:9300

Then restart Prometheus and you're good to go!

All of the metrics provided by cosmos-exporter have the following prefixes:

• cosmos_validator_* - metrics related to a single validator
• cosmos_validators_* - metrics related to a validator set
• cosmos_wallet_* - metrics related to a single wallet

How does it work?

It queries the full node via gRPC and returns it in the format Prometheus can consume.

How can I configure it?

You can pass the artuments to the executable file to configure it. Here is the parameters list:

• --bech-prefix - the global prefix for addresses. Defaults to persistence
• --denom - the currency, for example, uatom for Cosmos. Defaults to uxprt
• --listen-address - the address with port the node would listen to. For example, you can use it to redefine port or to make the exporter accessible from the outside by listening on 127.0.0.1. Defaults to :9300 (so it's accessible from the outside on port 9300)
• --node - the gRPC node URL. Defaults to localhost:9090
• --tendermint-rpc - Tendermint RPC URL to query node stats (specifically chain-id). Defaults to http://localhost:26657
• --log-devel - logger level. Defaults to info. You can set it to debug to make it more verbose.
• --limit - pagination limit for gRPC requests. Defaults to 1000.

You can also specify custom Bech32 prefixes for wallets, validators, consensus nodes, and their pubkeys by using the following params:

• --bech-account-prefix
• --bech-account-pubkey-prefix
• --bech-validator-prefix
• --bech-validator-pubkey-prefix
• --bech-consensus-node-prefix
• --bech-consensus-node-pubkey-prefix

By default, if not specified, it defaults to the next values (as it works this way for the most of the networks):

• --bech-account-prefix = --bech-prefix
• --bech-account-pubkey-prefix = --bech-prefix + "pub"
• --bech-validator-prefix = --bech-prefix + "valoper"
• --bech-validator-pubkey-prefix = --bech-prefix + "valoperpub"
• --bech-consensus-node-prefix = --bech-prefix + "valcons"
• --bech-consensus-node-pubkey-prefix = --bech-prefix + "valconspub"

An example of the network where you have to specify all the prefixes manually is Iris, check out the flags example below.

Additionally, you can pass a --config flag with a path to your config file (I use .toml, but anything supported by viper should work).

Which networks this is guaranteed to work?

In theory, it should work on a Cosmos-based blockchains that expose a gRPC endpoint (for example, Sentinel hub v0.5.0 doesn't expose it, so it won't work with it). In practice, this definitely works with the following blockchains:

• Persistence (with --bech-prefix persistence --denom uxprt)
• Cosmos (with --bech-prefix cosmos --denom uatom). The queries can take a long time though (apparently more than 90 seconds), so adjust the scrape timings and timeouts accordingly.
• Akash (with --bech-prefix akash --denom uakt)
• Iris (with --bech-account-prefix iaa --bech-consensus-node-prefix ica --bech-validator-prefix iva --bech-account-pubkey-prefix iap --bech-consensus-node-pubkey-prefix icp --bech-validator-pubkey-prefix ip --denom uiris)
• crypto.org (with --bech-account-prefix cro --bech-consensus-node-prefix crocnclcons --bech-validator-prefix crocncl --bech-account-pubkey-prefix cropub --bech-consensus-node-pubkey-prefix crocnclconspub --bech-validator-pubkey-prefix crocnclpub --denom basecro)

This won't work with the following blockchains:

• Sentinel - it doesn't expose gRPC endpoint for now. Should be fixed in sentinel-hub v0.6.0
• Kava - for the same reason.
• fetch.ai - for the same reason
• Terra - for the same reason
• Sifchain - for the same reason

How can I contribute?

Bug reports and feature requests are always welcome! If you want to contribute, feel free to open issues or PRs.