chore(docs): readme improvements

Includes:
- [x] Move the short path to the top
- [x] Include CLI use for the short path
- [x] Describe Mithril configuration
- [x] Explain use of nview

Signed-off-by: Chris Gianelloni <[email protected]>

Author: wolf31o2

README.md

@@ -18,10 +18,85 @@ invoked. There is a `NETWORK` environment variable which can be used as a
 short cut to starting nodes on a specific named Cardano network with a default
 configuration.
 
-### Using run (Recommended for SPO)
+We recommend using this method for running the container if you are not
+familiar with running a node.
+
+For nodes on the `preprod` and `mainnet` networks, the container will start
+fetching the latest Mithril snapshot if an empty data directory is detected.
+This snapshot will be used to bootstrap the node into an operating state much
+more rapidly than doing a sync from the beginning of the chain. This behavior
+can be disabled in the advanced path below.
+
+### Using NETWORK (Recommended)
+
+Using the `NETWORK` environment variable causes the entrypoint script to use
+the simplified path. The only configurable option in this path is the
+`NETWORK` environment variable itself.
+
+This is in keeping with the IOHK image behavior.
+
+To run a Cardano full node on preprod:
+
+```bash
+docker run --detach \
+  --name cardano-node \
+  -v node-data:/data/db \
+  -v node-ipc:/ipc \
+  -e NETWORK=preprod \
+  -p 3001:3001 \
+  ghcr.io/blinklabs-io/cardano-node
+```
+
+Node logs can be followed:
+
+```bash
+docker logs -f cardano-node
+```
+
+The node can be monitoring using [nview](https://github.com/blinklabs-io/nview)
+from within the container. We run `nview` from within the same container as
+our running node to get access to information about the node process.
+
+```bash
+docker exec -ti cardano-node nview
+```
+
+#### Running the Cardano CLI
+
+To run a Cardano CLI command, we use the `node-ipc` volume from our node.
+
+```bash
+docker run --rm -ti \
+  -e NETWORK=preprod \
+  -v node-ipc:/ipc \
+  ghcr.io/blinklabs-io/cardano-node cli <command>
+```
+
+This can be added to a shell alias:
+
+```bash
+alias cardano-cli="docker run --rm -ti \
+  -e NETWORK=preprod \
+  -v node-ipc:/ipc \
+  ghcr.io/blinklabs-io/cardano-node cli"
+```
+
+Now, you can use cardano-cli commands as you would, normally.
+
+```bash
+cardano-cli query tip --network-magic 1
+```
+
+Or, for a node running with NETWORK=mainnet:
+
+```bash
+cardano-cli query tip --mainnet
+```
+
+### Using run (Recommended for SPO/Advanced)
 
 Using the `run` argument to the image causes the entrypoint script to use the
-fully configurable code path.
+fully configurable code path. Do not set the `NETWORK` environment variable.
 
 To run a Cardano full node on mainnet with minimal configuration and defaults:
 
@@ -34,17 +109,53 @@ docker run --detach \
   ghcr.io/blinklabs-io/cardano-node run
 ```
 
+**NOTE** The container paths for persist disks are different in this mode
+
 The above maps port 3001 on the host to 3001 on the container, suitable
 for use as a mainnet relay node. We use docker volumes for our persistent data
 storage and IPC (interprocess communication) so they live beyond the lifetime
-of the container.
+of the container. To mount local host paths instead of using Docker's built in
+volumes, use the local path on the left side of a `-v` argument.
+
+An example of a more advanced configuration for mainnet:
+
+```bash
+docker run --detach \
+  --name cardano-node \
+  --restart unless-stopped \
+  -e CARDANO_CONFIG=/opt/cardano/config/mainnet/p2p-config.json \
+  -e CARDANO_TOPOLOGY=/opt/cardano/config/mainnet/p2p-topology.json \
+  -v /srv/cardano/node-db:/opt/cardano/data \
+  -v /srv/cardano/node-ipc:/opt/cardano/ipc \
+  -p 3001:3001 \
+  -p 12798:12798 \
+  ghcr.io/blinklabs-io/cardano-node run
+```
+
+The above uses Docker's built in supervisor to restart a container which fails
+for any reason. This will also cause the container to automatically restart
+after a host reboot, so long as Docker is configured to start on boot. The
+current default configuration for mainnet does not enable P2P, so we override
+the configuration with the shipped `p2p-config.json` and its accompanying
+`p2p-topology.json` to enable it. Our node's persistent data and client
+communication socket are mapped to `/src/cardano/node-db` and
+`/src/cardano/node-ipc` on the host, respectively. This allows for running
+applications directly on the host which may need access to these. Last, we add
+mapping the host's port 12798 to the container 12798, which is the port for
+exposing the node's metrics in Prometheus format, for monitoring.
+
+This mode of operation allows configuring multiple facets of the node using
+environment variables, described below.
 
 Node logs can be followed:
 
 ```bash
-docker logs -f cardano-node
+docker logs -f -n 500 cardano-node
 ```
 
+Adding the `-n 500` to the logs command limits the logs to the last 500 lines
+before following.
+
 #### Configuration using environment variables
 
 The power in using `run` is being able to configure the node's behavior to
@@ -99,29 +210,23 @@ and operational certificate.
   - Stake pool identity certificate (default:
     `${CARDANO_CONFIG_BASE}/keys/node.cert`
 
-### Using NETWORK (standalone)
+#### Controlling Mithril snapshots
 
-Using the `NETWORK` environment variable causes the entrypoint script to use
-the simplified path. The only configurable option in this path is the
-`NETWORK` environment variable itself.
+If the container does not find a protocolMagicId file within the
+`CARDANO_DATABASE_PATH` location, it will initiate Mithril snapshot downloads
+for preprod and mainnet networks. This can be disabled by setting
+`RESTORE_SNAPSHOT` to `false`.
 
-**NOTE** The container paths for persist disks is different in this mode
+- `AGGREGATOR_ENDPOINT`
+  - Mithril Aggregator URL (default:
+    `https://aggregator.release-${CARDANO_NETWORK}.api.mithril.network/aggregator`)
+- `GENESIS_VERIFICATION_KEY`
+  - Network specific Genesis Verification Key (default:
+    `reads file at: ${CARDANO_CONFIG_BASE}/${CARDANO_NETWORK}/genesis.vkey`)
+- `SNAPSHOT_DIGEST`
+  - Digest identifier to fetch (default: `latest`)
 
-This is in keeping with the IOHK image behavior.
-
-To run a Cardano full node on preview:
-
-```bash
-docker run --detach \
-  --name cardano-node \
-  -v node-data:/data/db \
-  -v node-ipc:/ipc \
-  -e NETWORK=preview \
-  -p 3001:3001 \
-  ghcr.io/blinklabs-io/cardano-node
-```
-
-## Running the CLI
+#### Running the Cardano CLI
 
 To run a Cardano CLI command, we use the `node-ipc` volume from our node.
 
@@ -144,9 +249,3 @@ Now, you can use cardano-cli commands as you would, normally.
 ```bash
 cardano-cli query tip --mainnet
 ```
-
-Or, for preview:
-
-```bash
-cardano-cli query tip --network-magic 2
-```