CLI and environment · CLIO Core

clio_run

usage

clio_run <command> [options]

Command	What it does
`start`	Start the runtime in the foreground, recover persistent state, and compose the pools in the config.
`stop`	Stop the runtime. `--grace-period <ms>` sets the drain budget (default 5000); `--force` stops immediately.
`status`	Print `RUNNING`, `STOPPED (clean)` or `STOPPED (stale artifacts)`. The exit code is 0, 1 or 2 to match.
`compose`	Manage pools from a YAML file: `start`, `stop`, `list`.
`monitor`	Print worker statistics.
`migrate`	Move a container to a different node.
`config export --path <file>`	Write the node's effective configuration to a file.
`refresh`	Regenerate method files for module development.

start options

Option	Effect
`--fresh`	Discard this node's persistent state instead of recovering it.
`--disk <path>`	Storage root for persistent state on this run (default `~/.clio`). Same as `CLIO_STORAGE_ROOT`.
`--ephemeral`	Skip the configured compose and start with only the admin pool.
`--induct`	Register this node with an existing cluster.
`--viz`, `--no-viz`	Serve or skip the web dashboard (on by default).
`--viz-port <port>`	Dashboard port (default 8080; 0 picks a free one).
`--viz-bind <addr>`	Dashboard bind address (default 127.0.0.1).

compose

A compose file has the same compose: list as clio.yaml. This one adds a 256 MB RAM block device:

yaml · tiers.yaml

compose:
  - mod_name: clio_bdev
    pool_name: "ram::scratch"
    pool_query: local
    pool_id: "320.0"
    bdev_type: ram
    capacity: "256MB"
    restart: true

shell

$ clio_run compose start tiers.yaml   # create the pools in the file
$ clio_run compose list               # list containers in the local daemon
$ clio_run compose stop tiers.yaml    # destroy those pools

Pools marked restart: true are recorded in each node's pool log and composed again by the next clio_run start. stop destroys the pools now but keeps them in the log. clio_run compose list --restartable lists the log.

clio_cte_fuse

usage

clio_cte_fuse <mountpoint> [FUSE options]

The daemon creates or binds the filesystem pool itself. Every argument after the mountpoint goes to FUSE: -f (foreground), -d (debug trace), -s (single thread), -o allow_other (needs user_allow_other in /etc/fuse.conf). On Windows the mountpoint is a drive letter (Z:) or an empty directory path. On macOS, -o backend=fskit selects FSKit.

Variable	Default	Effect
`CLIO_WITH_RUNTIME`	1	Set to `0` to attach to a running runtime. Almost always what you want.
`CLIO_FUSE_ATTR_CACHE_S`	1.0 (0 on several nodes)	Kernel attribute cache lifetime, in seconds. `0` gives strict coherence across clients.
`CLIO_FUSE_WRITEBACK`	off	Kernel writeback caching. Unsafe with several writers.

Environment variables

Variable	Effect
`CLIO_SERVER_CONF`	Path of the runtime configuration file (overrides `~/.clio/clio.yaml`).
`CLIO_STORAGE_ROOT`	Directory for persistent tiers, the metadata log and the index. Expanded as `${CLIO_STORAGE_ROOT}` in the config.
`CLIO_PORT`	Base RPC port (default 9413).
`CLIO_SERVER_ADDR`	Address clients use to reach the runtime.
`CLIO_IPC_MODE`	Client transport: `SHM` (shared memory), `IPC` (Unix socket) or `TCP`. When unset, the client picks the fastest one that works: SHM for a runtime on the same machine, otherwise IPC, otherwise TCP.
`CLIO_WITH_RUNTIME`	`0` attaches a client to a running runtime; `1` starts one in-process.
`CLIO_CTE_POOL`	Bind the CTE client to a chain layer, such as `564.0` for search.
`CLIO_VIZ_ENABLE`, `CLIO_VIZ_PORT`, `CLIO_VIZ_BIND`	Dashboard settings.
`CTP_LOG_LEVEL`	Log verbosity, such as `info` or `debug`.

Configuration file

~/.clio/clio.yaml has a few top-level sections. networking sets ports and, for a cluster, the hostfile with one node address per line. runtime sets worker threads and queue depth. viz configures the dashboard. compose lists the pools to create. The installed clio_default.yaml documents every key in comments. The compose tutorial covers the storage settings.