sleet 0.1.0

A fleet manager for SlateDB databases
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129

# CLI reference

Sleet has one long-running command, `run`, and several one-shot operator
commands for registration, status, and mirrors.

## Command overview

```text
sleet <COMMAND>

Commands:
  run       Run a fleet node
  status    Show fleet state and placement
  register  Register a database
  mirror    Run mirror operations
```

Use `-h` for a summary and `--help` for full help:

```sh
sleet run --help
sleet mirror restore --help
```

## `sleet run`

```sh
sleet run [OPTIONS] --node-id <NODE_ID> <ROOT>
```

Runs a node until interrupted.

Important options:

| Option                      | Meaning                                                   |
| --------------------------- | --------------------------------------------------------- |
| `--node-id <NODE_ID>`       | Required unique node identity.                            |
| `--services <SERVICES>`     | Comma-separated service list.                             |
| `--max-mirror-jobs <N>`     | Mirror copy or prune jobs on this node at once.           |
| `--rclone <PATH>`           | Binary for mirror destinations using `copier = "rclone"`. |

Service names:

```text
gc
compactor-coordinator
compaction-workers
mirror
```

Example:

```sh
sleet run s3://ops/sleet \
  --node-id worker-1 \
  --services compaction-workers
```

## `sleet register`

```sh
sleet register [OPTIONS] <ROOT> <DB>
```

Creates the registry file for a database. The operation is create-only, so it
does not overwrite operator edits.

```sh
sleet register s3://ops/sleet s3://bucket/db
sleet register s3://ops/sleet s3://bucket/db --format json
```

## `sleet status`

```sh
sleet status [OPTIONS] <ROOT>
```

Derives fleet state from the object-store tree.

| Option          | Meaning                                                  |
| --------------- | -------------------------------------------------------- |
| `--compactions` | Read each database's compaction queue depth.             |
| `--mirrors`     | Read mirror source and destination heads and report lag. |
| `--format json` | Emit JSON matching the CLI schema.                       |

Example:

```sh
sleet status s3://ops/sleet --compactions --mirrors
```

## `sleet mirror sync`

```sh
sleet mirror sync [OPTIONS] <ROOT> <DB> <TARGET>
```

Runs one sync pass for one registered database and destination. It prunes afterward
when retention is set.

```sh
sleet mirror sync s3://ops/sleet s3://bucket/db s3://backups/db1
```

Use `--rclone <PATH>` when the destination uses `copier = "rclone"`.

## `sleet mirror restore`

```sh
sleet mirror restore [OPTIONS] <BACKUP> <DEST>
```

Copies one restore point from a backup root into an empty destination root.

```sh
sleet mirror restore gs://backups/db1 s3://restore/db1
sleet mirror restore gs://backups/db1 s3://restore/db1 --at 42
```

`--at` accepts a manifest ID or RFC 3339 timestamp. A timestamp resolves to
the newest restore point at or before that time. The mapping uses the backup
manifest sequence tracker, which samples at about 60 seconds with the stock
SlateDB settings. If omitted, Sleet restores the backup's latest manifest.

## JSON outputs

One-shot commands that accept `--format json` emit responses defined in
[schema/cli.schema.json](../schema/cli.schema.json).