@bytesbrains/cruise 0.0.0-stage → 0.12.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/LICENSE +202 -0
- package/README.md +59 -2
- package/dist/budget-FIRNSN67.mjs +132 -0
- package/dist/chunk-2FB2T3UI.mjs +122 -0
- package/dist/chunk-2NVDAK5J.mjs +76 -0
- package/dist/chunk-42ERHAJD.mjs +79 -0
- package/dist/chunk-4447IUTF.mjs +7 -0
- package/dist/chunk-5TVD3ZAI.mjs +69 -0
- package/dist/chunk-7NQHBXVZ.mjs +104 -0
- package/dist/chunk-BNP63KB2.mjs +42 -0
- package/dist/chunk-FGLMW5IY.mjs +575 -0
- package/dist/chunk-GUZNI3NS.mjs +267 -0
- package/dist/chunk-HQDQTXRX.mjs +57 -0
- package/dist/chunk-ICJVJSTC.mjs +42 -0
- package/dist/chunk-JZPUTXM7.mjs +236 -0
- package/dist/chunk-LALZUKEC.mjs +137 -0
- package/dist/chunk-N3YKWKOA.mjs +9 -0
- package/dist/chunk-O7TNRKSA.mjs +2 -0
- package/dist/chunk-RT4R2W6H.mjs +21 -0
- package/dist/chunk-SYTYOTO4.mjs +18 -0
- package/dist/chunk-T3ICD43U.mjs +15 -0
- package/dist/chunk-THEXUHQC.mjs +98 -0
- package/dist/chunk-UNXM6FSM.mjs +83 -0
- package/dist/chunk-WS6NFRIU.mjs +66 -0
- package/dist/chunk-XOH3XKM3.mjs +126 -0
- package/dist/compress-3DWG35BY.mjs +49 -0
- package/dist/conformance-NPFDYCQL.mjs +922 -0
- package/dist/cruise.mjs +693 -0
- package/dist/fallback-P5KK55M4.mjs +139 -0
- package/dist/keys-FWPQPMTW.mjs +127 -0
- package/dist/ledger-XBGWXY6C.mjs +97 -0
- package/dist/member-ZBV4YG2M.mjs +74 -0
- package/dist/report-VRYJ3W5L.mjs +414 -0
- package/dist/service-key-N2RZ7FMZ.mjs +59 -0
- package/dist/smoke-ZNYX6RO5.mjs +105 -0
- package/dist/sweep-SZX7OIDK.mjs +33 -0
- package/dist/tenant-4YDGI23Q.mjs +122 -0
- package/dist/tier-4L3SPLFP.mjs +116 -0
- package/package.json +38 -4
- package/templates/docker-compose.yml +66 -0
- package/templates/env.example +44 -0
package/LICENSE
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
1
|
+
|
|
2
|
+
Apache License
|
|
3
|
+
Version 2.0, January 2004
|
|
4
|
+
http://www.apache.org/licenses/
|
|
5
|
+
|
|
6
|
+
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
|
7
|
+
|
|
8
|
+
1. Definitions.
|
|
9
|
+
|
|
10
|
+
"License" shall mean the terms and conditions for use, reproduction,
|
|
11
|
+
and distribution as defined by Sections 1 through 9 of this document.
|
|
12
|
+
|
|
13
|
+
"Licensor" shall mean the copyright owner or entity authorized by
|
|
14
|
+
the copyright owner that is granting the License.
|
|
15
|
+
|
|
16
|
+
"Legal Entity" shall mean the union of the acting entity and all
|
|
17
|
+
other entities that control, are controlled by, or are under common
|
|
18
|
+
control with that entity. For the purposes of this definition,
|
|
19
|
+
"control" means (i) the power, direct or indirect, to cause the
|
|
20
|
+
direction or management of such entity, whether by contract or
|
|
21
|
+
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
|
22
|
+
outstanding shares, or (iii) beneficial ownership of such entity.
|
|
23
|
+
|
|
24
|
+
"You" (or "Your") shall mean an individual or Legal Entity
|
|
25
|
+
exercising permissions granted by this License.
|
|
26
|
+
|
|
27
|
+
"Source" form shall mean the preferred form for making modifications,
|
|
28
|
+
including but not limited to software source code, documentation
|
|
29
|
+
source, and configuration files.
|
|
30
|
+
|
|
31
|
+
"Object" form shall mean any form resulting from mechanical
|
|
32
|
+
transformation or translation of a Source form, including but
|
|
33
|
+
not limited to compiled object code, generated documentation,
|
|
34
|
+
and conversions to other media types.
|
|
35
|
+
|
|
36
|
+
"Work" shall mean the work of authorship, whether in Source or
|
|
37
|
+
Object form, made available under the License, as indicated by a
|
|
38
|
+
copyright notice that is included in or attached to the work
|
|
39
|
+
(an example is provided in the Appendix below).
|
|
40
|
+
|
|
41
|
+
"Derivative Works" shall mean any work, whether in Source or Object
|
|
42
|
+
form, that is based on (or derived from) the Work and for which the
|
|
43
|
+
editorial revisions, annotations, elaborations, or other modifications
|
|
44
|
+
represent, as a whole, an original work of authorship. For the purposes
|
|
45
|
+
of this License, Derivative Works shall not include works that remain
|
|
46
|
+
separable from, or merely link (or bind by name) to the interfaces of,
|
|
47
|
+
the Work and Derivative Works thereof.
|
|
48
|
+
|
|
49
|
+
"Contribution" shall mean any work of authorship, including
|
|
50
|
+
the original version of the Work and any modifications or additions
|
|
51
|
+
to that Work or Derivative Works thereof, that is intentionally
|
|
52
|
+
submitted to Licensor for inclusion in the Work by the copyright owner
|
|
53
|
+
or by an individual or Legal Entity authorized to submit on behalf of
|
|
54
|
+
the copyright owner. For the purposes of this definition, "submitted"
|
|
55
|
+
means any form of electronic, verbal, or written communication sent
|
|
56
|
+
to the Licensor or its representatives, including but not limited to
|
|
57
|
+
communication on electronic mailing lists, source code control systems,
|
|
58
|
+
and issue tracking systems that are managed by, or on behalf of, the
|
|
59
|
+
Licensor for the purpose of discussing and improving the Work, but
|
|
60
|
+
excluding communication that is conspicuously marked or otherwise
|
|
61
|
+
designated in writing by the copyright owner as "Not a Contribution."
|
|
62
|
+
|
|
63
|
+
"Contributor" shall mean Licensor and any individual or Legal Entity
|
|
64
|
+
on behalf of whom a Contribution has been received by Licensor and
|
|
65
|
+
subsequently incorporated within the Work.
|
|
66
|
+
|
|
67
|
+
2. Grant of Copyright License. Subject to the terms and conditions of
|
|
68
|
+
this License, each Contributor hereby grants to You a perpetual,
|
|
69
|
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
|
70
|
+
copyright license to reproduce, prepare Derivative Works of,
|
|
71
|
+
publicly display, publicly perform, sublicense, and distribute the
|
|
72
|
+
Work and such Derivative Works in Source or Object form.
|
|
73
|
+
|
|
74
|
+
3. Grant of Patent License. Subject to the terms and conditions of
|
|
75
|
+
this License, each Contributor hereby grants to You a perpetual,
|
|
76
|
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
|
77
|
+
(except as stated in this section) patent license to make, have made,
|
|
78
|
+
use, offer to sell, sell, import, and otherwise transfer the Work,
|
|
79
|
+
where such license applies only to those patent claims licensable
|
|
80
|
+
by such Contributor that are necessarily infringed by their
|
|
81
|
+
Contribution(s) alone or by combination of their Contribution(s)
|
|
82
|
+
with the Work to which such Contribution(s) was submitted. If You
|
|
83
|
+
institute patent litigation against any entity (including a
|
|
84
|
+
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
|
85
|
+
or a Contribution incorporated within the Work constitutes direct
|
|
86
|
+
or contributory patent infringement, then any patent licenses
|
|
87
|
+
granted to You under this License for that Work shall terminate
|
|
88
|
+
as of the date such litigation is filed.
|
|
89
|
+
|
|
90
|
+
4. Redistribution. You may reproduce and distribute copies of the
|
|
91
|
+
Work or Derivative Works thereof in any medium, with or without
|
|
92
|
+
modifications, and in Source or Object form, provided that You
|
|
93
|
+
meet the following conditions:
|
|
94
|
+
|
|
95
|
+
(a) You must give any other recipients of the Work or
|
|
96
|
+
Derivative Works a copy of this License; and
|
|
97
|
+
|
|
98
|
+
(b) You must cause any modified files to carry prominent notices
|
|
99
|
+
stating that You changed the files; and
|
|
100
|
+
|
|
101
|
+
(c) You must retain, in the Source form of any Derivative Works
|
|
102
|
+
that You distribute, all copyright, patent, trademark, and
|
|
103
|
+
attribution notices from the Source form of the Work,
|
|
104
|
+
excluding those notices that do not pertain to any part of
|
|
105
|
+
the Derivative Works; and
|
|
106
|
+
|
|
107
|
+
(d) If the Work includes a "NOTICE" text file as part of its
|
|
108
|
+
distribution, then any Derivative Works that You distribute must
|
|
109
|
+
include a readable copy of the attribution notices contained
|
|
110
|
+
within such NOTICE file, excluding those notices that do not
|
|
111
|
+
pertain to any part of the Derivative Works, in at least one
|
|
112
|
+
of the following places: within a NOTICE text file distributed
|
|
113
|
+
as part of the Derivative Works; within the Source form or
|
|
114
|
+
documentation, if provided along with the Derivative Works; or,
|
|
115
|
+
within a display generated by the Derivative Works, if and
|
|
116
|
+
wherever such third-party notices normally appear. The contents
|
|
117
|
+
of the NOTICE file are for informational purposes only and
|
|
118
|
+
do not modify the License. You may add Your own attribution
|
|
119
|
+
notices within Derivative Works that You distribute, alongside
|
|
120
|
+
or as an addendum to the NOTICE text from the Work, provided
|
|
121
|
+
that such additional attribution notices cannot be construed
|
|
122
|
+
as modifying the License.
|
|
123
|
+
|
|
124
|
+
You may add Your own copyright statement to Your modifications and
|
|
125
|
+
may provide additional or different license terms and conditions
|
|
126
|
+
for use, reproduction, or distribution of Your modifications, or
|
|
127
|
+
for any such Derivative Works as a whole, provided Your use,
|
|
128
|
+
reproduction, and distribution of the Work otherwise complies with
|
|
129
|
+
the conditions stated in this License.
|
|
130
|
+
|
|
131
|
+
5. Submission of Contributions. Unless You explicitly state otherwise,
|
|
132
|
+
any Contribution intentionally submitted for inclusion in the Work
|
|
133
|
+
by You to the Licensor shall be under the terms and conditions of
|
|
134
|
+
this License, without any additional terms or conditions.
|
|
135
|
+
Notwithstanding the above, nothing herein shall supersede or modify
|
|
136
|
+
the terms of any separate license agreement you may have executed
|
|
137
|
+
with Licensor regarding such Contributions.
|
|
138
|
+
|
|
139
|
+
6. Trademarks. This License does not grant permission to use the trade
|
|
140
|
+
names, trademarks, service marks, or product names of the Licensor,
|
|
141
|
+
except as required for reasonable and customary use in describing the
|
|
142
|
+
origin of the Work and reproducing the content of the NOTICE file.
|
|
143
|
+
|
|
144
|
+
7. Disclaimer of Warranty. Unless required by applicable law or
|
|
145
|
+
agreed to in writing, Licensor provides the Work (and each
|
|
146
|
+
Contributor provides its Contributions) on an "AS IS" BASIS,
|
|
147
|
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
|
148
|
+
implied, including, without limitation, any warranties or conditions
|
|
149
|
+
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
|
150
|
+
PARTICULAR PURPOSE. You are solely responsible for determining the
|
|
151
|
+
appropriateness of using or redistributing the Work and assume any
|
|
152
|
+
risks associated with Your exercise of permissions under this License.
|
|
153
|
+
|
|
154
|
+
8. Limitation of Liability. In no event and under no legal theory,
|
|
155
|
+
whether in tort (including negligence), contract, or otherwise,
|
|
156
|
+
unless required by applicable law (such as deliberate and grossly
|
|
157
|
+
negligent acts) or agreed to in writing, shall any Contributor be
|
|
158
|
+
liable to You for damages, including any direct, indirect, special,
|
|
159
|
+
incidental, or consequential damages of any character arising as a
|
|
160
|
+
result of this License or out of the use or inability to use the
|
|
161
|
+
Work (including but not limited to damages for loss of goodwill,
|
|
162
|
+
work stoppage, computer failure or malfunction, or any and all
|
|
163
|
+
other commercial damages or losses), even if such Contributor
|
|
164
|
+
has been advised of the possibility of such damages.
|
|
165
|
+
|
|
166
|
+
9. Accepting Warranty or Additional Liability. While redistributing
|
|
167
|
+
the Work or Derivative Works thereof, You may choose to offer,
|
|
168
|
+
and charge a fee for, acceptance of support, warranty, indemnity,
|
|
169
|
+
or other liability obligations and/or rights consistent with this
|
|
170
|
+
License. However, in accepting such obligations, You may act only
|
|
171
|
+
on Your own behalf and on Your sole responsibility, not on behalf
|
|
172
|
+
of any other Contributor, and only if You agree to indemnify,
|
|
173
|
+
defend, and hold each Contributor harmless for any liability
|
|
174
|
+
incurred by, or claims asserted against, such Contributor by reason
|
|
175
|
+
of your accepting any such warranty or additional liability.
|
|
176
|
+
|
|
177
|
+
END OF TERMS AND CONDITIONS
|
|
178
|
+
|
|
179
|
+
APPENDIX: How to apply the Apache License to your work.
|
|
180
|
+
|
|
181
|
+
To apply the Apache License to your work, attach the following
|
|
182
|
+
boilerplate notice, with the fields enclosed by brackets "[]"
|
|
183
|
+
replaced with your own identifying information. (Don't include
|
|
184
|
+
the brackets!) The text should be enclosed in the appropriate
|
|
185
|
+
comment syntax for the file format. We also recommend that a
|
|
186
|
+
file or class name and description of purpose be included on the
|
|
187
|
+
same "printed page" as the copyright notice for easier
|
|
188
|
+
identification within third-party archives.
|
|
189
|
+
|
|
190
|
+
Copyright [yyyy] [name of copyright owner]
|
|
191
|
+
|
|
192
|
+
Licensed under the Apache License, Version 2.0 (the "License");
|
|
193
|
+
you may not use this file except in compliance with the License.
|
|
194
|
+
You may obtain a copy of the License at
|
|
195
|
+
|
|
196
|
+
http://www.apache.org/licenses/LICENSE-2.0
|
|
197
|
+
|
|
198
|
+
Unless required by applicable law or agreed to in writing, software
|
|
199
|
+
distributed under the License is distributed on an "AS IS" BASIS,
|
|
200
|
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
201
|
+
See the License for the specific language governing permissions and
|
|
202
|
+
limitations under the License.
|
package/README.md
CHANGED
|
@@ -1,3 +1,60 @@
|
|
|
1
|
-
#
|
|
1
|
+
# cruise
|
|
2
2
|
|
|
3
|
-
|
|
3
|
+
**One OpenAI-compatible endpoint in front of every model provider**, with per-project keys,
|
|
4
|
+
budgets that stop a runaway loop, a cost ledger every request lands in, and lanes that pick a
|
|
5
|
+
model per request. This package installs and runs [Cruise](https://gitlab.com/bytesbrains-oss/cruise),
|
|
6
|
+
in Docker on your machine or on your own Cloudflare account.
|
|
7
|
+
|
|
8
|
+
Node 22 or later.
|
|
9
|
+
|
|
10
|
+
## In Docker
|
|
11
|
+
|
|
12
|
+
```sh
|
|
13
|
+
npx @bytesbrains/cruise up
|
|
14
|
+
```
|
|
15
|
+
|
|
16
|
+
On first run this asks for your provider keys (input hidden; skip any), writes
|
|
17
|
+
`~/.cruise/docker/`, starts the container, waits until it is healthy, and prints your first API
|
|
18
|
+
key, **once**. Then point any OpenAI SDK at `http://localhost:8787/v1`.
|
|
19
|
+
|
|
20
|
+
Before a model answers, Cruise has to have seen it serve. Measure every model you hold a key
|
|
21
|
+
for, and again weekly:
|
|
22
|
+
|
|
23
|
+
```sh
|
|
24
|
+
npx @bytesbrains/cruise measure
|
|
25
|
+
```
|
|
26
|
+
|
|
27
|
+
| | |
|
|
28
|
+
|---|---|
|
|
29
|
+
| `cruise init` | Write `~/.cruise/docker/` and ask for keys, without starting |
|
|
30
|
+
| `cruise up` | Start, or upgrade the running container to new settings |
|
|
31
|
+
| `cruise status`, `cruise logs [-f]` | What is running, and what it said |
|
|
32
|
+
| `cruise backup` | A consistent backup now, into the data volume |
|
|
33
|
+
| `cruise upgrade` | Back up, pull the newest image, restart. The image migrates before it serves |
|
|
34
|
+
| `cruise down` | Stop. The data volume is kept |
|
|
35
|
+
| `cruise measure` | Measure every model you hold a key for |
|
|
36
|
+
| `cruise key:issue`, `budget:set`, `ledger`, `report`, … | Admin, inside the container. `cruise help` lists them |
|
|
37
|
+
|
|
38
|
+
Keys go in `~/.cruise/docker/.env` (mode 0600). Run `cruise up` again after editing it.
|
|
39
|
+
`CRUISE_HOME` moves the whole directory.
|
|
40
|
+
|
|
41
|
+
## On your own Cloudflare account
|
|
42
|
+
|
|
43
|
+
```sh
|
|
44
|
+
export CLOUDFLARE_API_TOKEN=… # Workers Scripts, D1, Workers KV Storage and Workers R2 Storage edit; Account Settings read
|
|
45
|
+
npx @bytesbrains/cruise cloudflare deploy --env-file .env
|
|
46
|
+
npx @bytesbrains/cruise cloudflare measure
|
|
47
|
+
```
|
|
48
|
+
|
|
49
|
+
`deploy` creates the database, cache and bucket, applies the migrations, deploys the Worker,
|
|
50
|
+
sets the secrets and the `PROVIDER_KEY_*` values from `--env-file`, and creates your first tenant
|
|
51
|
+
and key. Run it again to repair a deploy that stopped partway or to update to this package's
|
|
52
|
+
version; it never duplicates anything. `--name` deploys more than one.
|
|
53
|
+
|
|
54
|
+
Every provider is called at its own API, so Workers AI and AI Gateway are not used. Admin
|
|
55
|
+
commands work the same way: `cruise cloudflare ledger --by model`.
|
|
56
|
+
|
|
57
|
+
## More
|
|
58
|
+
|
|
59
|
+
The manual is [`docs/self-hosting.md`](https://gitlab.com/bytesbrains-oss/cruise/-/blob/main/docs/self-hosting.md).
|
|
60
|
+
Apache-2.0.
|
|
@@ -0,0 +1,132 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
// Built by packages/cli/build.ts.
|
|
3
|
+
import {
|
|
4
|
+
DEMO_ENVIRONMENT
|
|
5
|
+
} from "./chunk-N3YKWKOA.mjs";
|
|
6
|
+
import {
|
|
7
|
+
demoChargeStatements
|
|
8
|
+
} from "./chunk-RT4R2W6H.mjs";
|
|
9
|
+
import {
|
|
10
|
+
formatUsd,
|
|
11
|
+
readBudget,
|
|
12
|
+
setBudget,
|
|
13
|
+
usdToMicros
|
|
14
|
+
} from "./chunk-THEXUHQC.mjs";
|
|
15
|
+
import "./chunk-LALZUKEC.mjs";
|
|
16
|
+
import {
|
|
17
|
+
connect,
|
|
18
|
+
environmentOf,
|
|
19
|
+
optionalNumber,
|
|
20
|
+
parseArgs,
|
|
21
|
+
requireString,
|
|
22
|
+
run
|
|
23
|
+
} from "./chunk-7NQHBXVZ.mjs";
|
|
24
|
+
import "./chunk-JZPUTXM7.mjs";
|
|
25
|
+
|
|
26
|
+
// scripts/budget.ts
|
|
27
|
+
run(async () => {
|
|
28
|
+
const [command, ...rest] = process.argv.slice(2);
|
|
29
|
+
const args = parseArgs(rest);
|
|
30
|
+
const environment = environmentOf(args);
|
|
31
|
+
const db = connect(environment);
|
|
32
|
+
switch (command) {
|
|
33
|
+
case "set": {
|
|
34
|
+
const project = requireString(args, "project");
|
|
35
|
+
const period = typeof args.period === "string" ? args.period : "day";
|
|
36
|
+
const demoOnly = environment === DEMO_ENVIRONMENT;
|
|
37
|
+
if (period === "minute" && !demoOnly) throw new Error("--period minute is for --env demo only (#494)");
|
|
38
|
+
if (period !== "minute" && period !== "day" && period !== "month") {
|
|
39
|
+
throw new Error("--period must be day or month (or minute, on the demo)");
|
|
40
|
+
}
|
|
41
|
+
const charge = optionalNumber(args, "charge");
|
|
42
|
+
if (charge !== null && !demoOnly) throw new Error("--charge is for --env demo only (#494)");
|
|
43
|
+
const soft = optionalNumber(args, "soft");
|
|
44
|
+
const hard = optionalNumber(args, "hard");
|
|
45
|
+
const actor = requireString(args, "actor");
|
|
46
|
+
const sessionLimit = args["session-limit"] === "none" ? null : optionalNumber(args, "session-limit");
|
|
47
|
+
if (sessionLimit !== null && sessionLimit < 0) throw new Error("--session-limit must be a non-negative number, or none");
|
|
48
|
+
await setBudget(
|
|
49
|
+
db,
|
|
50
|
+
project,
|
|
51
|
+
{
|
|
52
|
+
period,
|
|
53
|
+
softMicros: soft === null ? null : usdToMicros(soft),
|
|
54
|
+
hardMicros: hard === null ? null : usdToMicros(hard),
|
|
55
|
+
// Absent leaves the per-run cap alone; `none` is the explicit clear.
|
|
56
|
+
...args["session-limit"] === void 0 ? {} : { sessionHardMicros: sessionLimit === null ? null : usdToMicros(sessionLimit) }
|
|
57
|
+
},
|
|
58
|
+
actor,
|
|
59
|
+
charge === null ? [] : demoChargeStatements(db, project, usdToMicros(charge), actor)
|
|
60
|
+
);
|
|
61
|
+
const status = await readBudget(db, project);
|
|
62
|
+
const softLabel = status?.softMicros == null ? "\u2014" : formatUsd(status.softMicros);
|
|
63
|
+
const hardLabel = status?.hardMicros == null ? "UNCAPPED" : formatUsd(status.hardMicros);
|
|
64
|
+
const runLabel = status?.sessionHardMicros == null ? "\u2014" : formatUsd(status.sessionHardMicros);
|
|
65
|
+
console.log(`${project}: soft ${softLabel} hard ${hardLabel} per ${status?.period} per run ${runLabel}`);
|
|
66
|
+
if (charge !== null) console.log(` demo charge ${formatUsd(usdToMicros(charge), 4)} per served request; reservation floor 0`);
|
|
67
|
+
if (status?.hardMicros == null) {
|
|
68
|
+
console.log(" warning: no hard cap. This is the condition that keeps the pipelines switched off.");
|
|
69
|
+
}
|
|
70
|
+
break;
|
|
71
|
+
}
|
|
72
|
+
case "show": {
|
|
73
|
+
const { results } = await db.prepare(`SELECT id FROM projects ORDER BY id`).all();
|
|
74
|
+
console.log("project state spend soft hard at cap");
|
|
75
|
+
for (const { id } of results) {
|
|
76
|
+
const s = await readBudget(db, id);
|
|
77
|
+
if (!s) continue;
|
|
78
|
+
console.log(
|
|
79
|
+
`${id.padEnd(9)} ${s.state.padEnd(7)} ${formatUsd(s.spendMicros, 4).padStart(8)} ${(s.softMicros == null ? "\u2014" : formatUsd(s.softMicros)).padStart(6)} ${(s.hardMicros == null ? "NONE" : formatUsd(s.hardMicros)).padStart(6)} ${s.behaviour}`
|
|
80
|
+
);
|
|
81
|
+
}
|
|
82
|
+
break;
|
|
83
|
+
}
|
|
84
|
+
case "align": {
|
|
85
|
+
const project = requireString(args, "project");
|
|
86
|
+
const day = requireString(args, "day");
|
|
87
|
+
if (!/^\d{4}-\d{2}-\d{2}$/.test(day)) throw new Error("--day must be YYYY-MM-DD");
|
|
88
|
+
const actor = requireString(args, "actor");
|
|
89
|
+
const apply = args.apply === true || args.apply === "true";
|
|
90
|
+
const counter = await db.prepare(
|
|
91
|
+
`SELECT spend_micros FROM spend
|
|
92
|
+
WHERE scope = 'project' AND subject = ?1 AND period = 'day' AND period_start = ?2`
|
|
93
|
+
).bind(project, day).first();
|
|
94
|
+
const ledger = await db.prepare(
|
|
95
|
+
`SELECT COALESCE(sum(cost_micros), 0) AS charged FROM ledger
|
|
96
|
+
WHERE project_id = ?1 AND date(at) = ?2`
|
|
97
|
+
).bind(project, day).first();
|
|
98
|
+
const before = counter?.spend_micros ?? 0;
|
|
99
|
+
const charged = ledger?.charged ?? 0;
|
|
100
|
+
const delta = charged - before;
|
|
101
|
+
console.log(
|
|
102
|
+
`${project} day ${day}: counter ${before} ledger ${charged} delta ${delta}` + (delta === 0 ? " (already aligned)" : "")
|
|
103
|
+
);
|
|
104
|
+
if (delta === 0 || !apply) {
|
|
105
|
+
if (delta !== 0 && !apply) console.log("dry-run; pass --apply to write");
|
|
106
|
+
break;
|
|
107
|
+
}
|
|
108
|
+
await db.batch([
|
|
109
|
+
db.prepare(
|
|
110
|
+
`INSERT INTO spend (scope, subject, period, period_start, spend_micros, request_count)
|
|
111
|
+
VALUES ('project', ?1, 'day', ?2, ?3, 0)
|
|
112
|
+
ON CONFLICT (scope, subject, period, period_start) DO UPDATE
|
|
113
|
+
SET spend_micros = ?3, updated_at = datetime('now')`
|
|
114
|
+
).bind(project, day, charged),
|
|
115
|
+
db.prepare(
|
|
116
|
+
`INSERT INTO audit_events (id, actor, action, subject, detail, tenant_id)
|
|
117
|
+
SELECT ?1, ?2, 'budget.align', ?3, ?4, (SELECT tenant_id FROM projects WHERE id = ?3)`
|
|
118
|
+
).bind(
|
|
119
|
+
crypto.randomUUID(),
|
|
120
|
+
actor,
|
|
121
|
+
project,
|
|
122
|
+
JSON.stringify({ day, before_micros: before, after_micros: charged, delta_micros: delta })
|
|
123
|
+
)
|
|
124
|
+
]);
|
|
125
|
+
console.log(`aligned to ${charged}`);
|
|
126
|
+
break;
|
|
127
|
+
}
|
|
128
|
+
default:
|
|
129
|
+
console.error("usage: show | set | align (see scripts/budget.ts)");
|
|
130
|
+
process.exitCode = 1;
|
|
131
|
+
}
|
|
132
|
+
});
|
|
@@ -0,0 +1,122 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
// Built by packages/cli/build.ts.
|
|
3
|
+
import {
|
|
4
|
+
DEFAULT_CACHE_POOL
|
|
5
|
+
} from "./chunk-T3ICD43U.mjs";
|
|
6
|
+
import {
|
|
7
|
+
generateKey,
|
|
8
|
+
keyRateLimitFor
|
|
9
|
+
} from "./chunk-5TVD3ZAI.mjs";
|
|
10
|
+
|
|
11
|
+
// src/admin.ts
|
|
12
|
+
function scopeProblem(entries, what) {
|
|
13
|
+
if (!Array.isArray(entries)) return `'${what}' must be an array of strings.`;
|
|
14
|
+
if (entries.some((entry) => typeof entry !== "string" || entry.trim() === "")) {
|
|
15
|
+
return `'${what}' has an empty entry. An unset variable reads exactly like this, and a blank is not a scope: omit the field for an unrestricted key.`;
|
|
16
|
+
}
|
|
17
|
+
return null;
|
|
18
|
+
}
|
|
19
|
+
function rateLimitProblem(value, what) {
|
|
20
|
+
if (value === void 0 || value === null) return null;
|
|
21
|
+
if (typeof value !== "number" || !Number.isInteger(value) || value < 1) {
|
|
22
|
+
return `'${what}' must be a positive integer. Omit the field, or send null, for no limit on that axis.`;
|
|
23
|
+
}
|
|
24
|
+
return null;
|
|
25
|
+
}
|
|
26
|
+
async function platformPoolProblem(db, projectId) {
|
|
27
|
+
const row = await db.prepare(
|
|
28
|
+
`SELECT t.operator FROM projects p JOIN tenants t ON t.id = p.tenant_id WHERE p.id = ?`
|
|
29
|
+
).bind(projectId).first();
|
|
30
|
+
if (row?.operator === 1) return null;
|
|
31
|
+
return `'${projectId}' does not belong to the tenant that operates Cruise, so a key for it cannot carry the 'platform' cache pool: that pool is read by every project we operate. Use 'tenant' to share within this tenant.`;
|
|
32
|
+
}
|
|
33
|
+
function canonicalScope(entries, what) {
|
|
34
|
+
const problem = scopeProblem(entries ?? [], what);
|
|
35
|
+
if (problem !== null) throw new Error(problem);
|
|
36
|
+
return [...new Set((entries ?? []).map((entry) => entry.trim()))];
|
|
37
|
+
}
|
|
38
|
+
async function issueKey(db, request) {
|
|
39
|
+
const allowedModels = canonicalScope(request.allowedModels, "models");
|
|
40
|
+
const allowedClasses = canonicalScope(request.allowedClasses, "classes");
|
|
41
|
+
const project = await db.prepare(`SELECT id FROM projects WHERE id = ? AND tenant_id = ?`).bind(request.projectId, request.tenantId).first();
|
|
42
|
+
if (project === null) throw new Error(`No such project: ${request.projectId}`);
|
|
43
|
+
const cachePool = request.cachePool ?? DEFAULT_CACHE_POOL;
|
|
44
|
+
if (cachePool === "platform") {
|
|
45
|
+
const problem = await platformPoolProblem(db, project.id);
|
|
46
|
+
if (problem !== null) throw new Error(problem);
|
|
47
|
+
}
|
|
48
|
+
const generated = await generateKey(request.environment);
|
|
49
|
+
const keyId = crypto.randomUUID();
|
|
50
|
+
const rateLimitPerMin = keyRateLimitFor(request.environment, request.rateLimitPerMin);
|
|
51
|
+
const accountRateLimitPerMin = request.accountRateLimitPerMin ?? null;
|
|
52
|
+
const statements = [
|
|
53
|
+
db.prepare(
|
|
54
|
+
`INSERT INTO virtual_keys
|
|
55
|
+
(id, project_id, name, prefix, key_hash, created_by,
|
|
56
|
+
rate_limit_per_min, account_rate_limit_per_min,
|
|
57
|
+
conformance, cache_pool, revoked_at, revoked_reason)
|
|
58
|
+
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, datetime('now'), 'provisioning')`
|
|
59
|
+
).bind(
|
|
60
|
+
keyId,
|
|
61
|
+
project.id,
|
|
62
|
+
request.name ?? null,
|
|
63
|
+
generated.prefix,
|
|
64
|
+
generated.hash,
|
|
65
|
+
request.actor,
|
|
66
|
+
rateLimitPerMin,
|
|
67
|
+
accountRateLimitPerMin,
|
|
68
|
+
request.conformance ? 1 : 0,
|
|
69
|
+
cachePool
|
|
70
|
+
)
|
|
71
|
+
];
|
|
72
|
+
for (const model of allowedModels) {
|
|
73
|
+
statements.push(
|
|
74
|
+
db.prepare(`INSERT INTO key_allowed_models (key_id, model) VALUES (?, ?)`).bind(keyId, model)
|
|
75
|
+
);
|
|
76
|
+
}
|
|
77
|
+
for (const trafficClass of allowedClasses) {
|
|
78
|
+
statements.push(
|
|
79
|
+
db.prepare(`INSERT INTO key_allowed_classes (key_id, class) VALUES (?, ?)`).bind(keyId, trafficClass)
|
|
80
|
+
);
|
|
81
|
+
}
|
|
82
|
+
statements.push(
|
|
83
|
+
db.prepare(
|
|
84
|
+
`INSERT INTO audit_events (id, actor, action, subject, detail, tenant_id)
|
|
85
|
+
SELECT ?1, ?2, 'key.issue', ?3, ?4, (SELECT tenant_id FROM projects WHERE id = ?5)`
|
|
86
|
+
).bind(
|
|
87
|
+
crypto.randomUUID(),
|
|
88
|
+
request.actor,
|
|
89
|
+
generated.prefix,
|
|
90
|
+
JSON.stringify({ project: project.id, cache_pool: cachePool }),
|
|
91
|
+
project.id
|
|
92
|
+
)
|
|
93
|
+
);
|
|
94
|
+
statements.push(
|
|
95
|
+
db.prepare(`UPDATE virtual_keys SET revoked_at = NULL, revoked_reason = NULL WHERE id = ?`).bind(keyId)
|
|
96
|
+
);
|
|
97
|
+
await db.batch(statements);
|
|
98
|
+
return { keyId, prefix: generated.prefix, key: generated.key, rateLimitPerMin, accountRateLimitPerMin };
|
|
99
|
+
}
|
|
100
|
+
async function revokeKey(db, request) {
|
|
101
|
+
const { tenantId, prefix, actor, reason } = request;
|
|
102
|
+
const result = await db.prepare(
|
|
103
|
+
`UPDATE virtual_keys
|
|
104
|
+
SET revoked_at = datetime('now'), revoked_by = ?, revoked_reason = ?
|
|
105
|
+
WHERE prefix = ? AND revoked_at IS NULL
|
|
106
|
+
AND project_id IN (SELECT id FROM projects WHERE tenant_id = ?)`
|
|
107
|
+
).bind(actor, reason, prefix, tenantId).run();
|
|
108
|
+
const changed = (result.meta.changes ?? 0) > 0;
|
|
109
|
+
if (changed) {
|
|
110
|
+
await db.prepare(
|
|
111
|
+
`INSERT INTO audit_events (id, actor, action, subject, reason, tenant_id)
|
|
112
|
+
VALUES (?, ?, 'key.revoke', ?, ?, ?)`
|
|
113
|
+
).bind(crypto.randomUUID(), actor, prefix, reason, tenantId).run();
|
|
114
|
+
}
|
|
115
|
+
return changed;
|
|
116
|
+
}
|
|
117
|
+
|
|
118
|
+
export {
|
|
119
|
+
rateLimitProblem,
|
|
120
|
+
issueKey,
|
|
121
|
+
revokeKey
|
|
122
|
+
};
|
|
@@ -0,0 +1,76 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
// Built by packages/cli/build.ts.
|
|
3
|
+
|
|
4
|
+
// src/membership.ts
|
|
5
|
+
var ROLES = ["owner", "admin", "viewer"];
|
|
6
|
+
var MAX_UID_LENGTH = 128;
|
|
7
|
+
function assertUid(uid) {
|
|
8
|
+
const trimmed = uid.trim();
|
|
9
|
+
if (trimmed === "") throw new Error("uid cannot be empty");
|
|
10
|
+
if (trimmed.length > MAX_UID_LENGTH) {
|
|
11
|
+
throw new Error(`uid cannot exceed ${MAX_UID_LENGTH} characters`);
|
|
12
|
+
}
|
|
13
|
+
return trimmed;
|
|
14
|
+
}
|
|
15
|
+
async function grantMembership(db, rawUid, tenantId, role, actor) {
|
|
16
|
+
const uid = assertUid(rawUid);
|
|
17
|
+
if (!ROLES.includes(role)) throw new Error(`Unknown role: ${role}`);
|
|
18
|
+
const tenant = await db.prepare(`SELECT id FROM tenants WHERE id = ?`).bind(tenantId).first();
|
|
19
|
+
if (tenant === null) throw new Error(`No such tenant: ${tenantId}`);
|
|
20
|
+
await db.batch([
|
|
21
|
+
// Re-granting changes the role rather than failing, because "make Alice an
|
|
22
|
+
// admin" and "Alice is already a viewer" should not be two different
|
|
23
|
+
// operations for whoever is doing it.
|
|
24
|
+
db.prepare(
|
|
25
|
+
`INSERT INTO memberships (uid, tenant_id, role, created_by) VALUES (?1, ?2, ?3, ?4)
|
|
26
|
+
ON CONFLICT (uid, tenant_id) DO UPDATE SET role = ?3, created_by = ?4`
|
|
27
|
+
).bind(uid, tenantId, role, actor),
|
|
28
|
+
db.prepare(
|
|
29
|
+
`INSERT INTO audit_events (id, actor, action, subject, detail, tenant_id)
|
|
30
|
+
VALUES (?, ?, 'membership.grant', ?, ?, ?)`
|
|
31
|
+
).bind(crypto.randomUUID(), actor, uid, JSON.stringify({ tenant: tenantId, role }), tenantId)
|
|
32
|
+
]);
|
|
33
|
+
}
|
|
34
|
+
async function revokeMembership(db, uid, tenantId, actor, reason) {
|
|
35
|
+
const result = await db.prepare(
|
|
36
|
+
`DELETE FROM memberships
|
|
37
|
+
WHERE uid = ?1 AND tenant_id = ?2
|
|
38
|
+
AND NOT (
|
|
39
|
+
role = 'owner'
|
|
40
|
+
AND (SELECT count(*) FROM memberships WHERE tenant_id = ?2 AND role = 'owner') <= 1
|
|
41
|
+
)`
|
|
42
|
+
).bind(uid, tenantId).run();
|
|
43
|
+
if ((result.meta.changes ?? 0) > 0) {
|
|
44
|
+
await db.prepare(
|
|
45
|
+
`INSERT INTO audit_events (id, actor, action, subject, reason, tenant_id)
|
|
46
|
+
VALUES (?, ?, 'membership.revoke', ?, ?, ?)`
|
|
47
|
+
).bind(crypto.randomUUID(), actor, uid, reason, tenantId).run();
|
|
48
|
+
return "revoked";
|
|
49
|
+
}
|
|
50
|
+
const still = await db.prepare(`SELECT 1 AS ok FROM memberships WHERE uid = ? AND tenant_id = ?`).bind(uid, tenantId).first();
|
|
51
|
+
return still === null ? "not_found" : "last_owner";
|
|
52
|
+
}
|
|
53
|
+
async function listMemberships(db, tenantId) {
|
|
54
|
+
const statement = tenantId ? db.prepare(
|
|
55
|
+
`SELECT uid, tenant_id, role, created_at, created_by FROM memberships
|
|
56
|
+
WHERE tenant_id = ? ORDER BY created_at`
|
|
57
|
+
).bind(tenantId) : db.prepare(
|
|
58
|
+
`SELECT uid, tenant_id, role, created_at, created_by FROM memberships ORDER BY tenant_id, created_at`
|
|
59
|
+
);
|
|
60
|
+
const { results } = await statement.all();
|
|
61
|
+
return results.map((row) => ({
|
|
62
|
+
uid: row.uid,
|
|
63
|
+
tenantId: row.tenant_id,
|
|
64
|
+
role: row.role,
|
|
65
|
+
createdAt: row.created_at,
|
|
66
|
+
createdBy: row.created_by
|
|
67
|
+
}));
|
|
68
|
+
}
|
|
69
|
+
|
|
70
|
+
export {
|
|
71
|
+
ROLES,
|
|
72
|
+
assertUid,
|
|
73
|
+
grantMembership,
|
|
74
|
+
revokeMembership,
|
|
75
|
+
listMemberships
|
|
76
|
+
};
|
|
@@ -0,0 +1,79 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
// Built by packages/cli/build.ts.
|
|
3
|
+
|
|
4
|
+
// src/ledger-lanes.ts
|
|
5
|
+
async function laneSavings(db, month, tenantId, laneId) {
|
|
6
|
+
const scope = "AND (?2 IS NULL OR l.tenant_id = ?2) AND (?3 IS NULL OR l.lane = ?3)";
|
|
7
|
+
const statement = db.prepare(
|
|
8
|
+
`WITH lane_requests AS (
|
|
9
|
+
SELECT l.id, l.lane, l.model, l.cost_micros, l.list_micros, l.cache_status
|
|
10
|
+
FROM ledger l
|
|
11
|
+
WHERE (?1 IS NULL OR strftime('%Y-%m', l.at) = ?1) AND l.priced = 1
|
|
12
|
+
AND l.requested_model = l.lane
|
|
13
|
+
-- A lane request is one whose requested model IS a lane, checked
|
|
14
|
+
-- against the table rather than by prefix. This read
|
|
15
|
+
-- 'l.lane LIKE bb/%', written when every lane was ours and left there
|
|
16
|
+
-- when 0028 gave tenants their own -- so a tenant lane's savings, the
|
|
17
|
+
-- receipt that is the entire reason to put config in a lane, came
|
|
18
|
+
-- back empty. The lane column is caller-influenced (x-cruise-lane),
|
|
19
|
+
-- so the existence check also keeps a spoofed header out.
|
|
20
|
+
AND EXISTS (SELECT 1 FROM lanes ln WHERE ln.id = l.lane) ${scope}
|
|
21
|
+
),
|
|
22
|
+
dearest AS (
|
|
23
|
+
SELECT a.ledger_id, max(a.cost_micros) AS cost
|
|
24
|
+
FROM lane_alternatives a WHERE a.priced = 1 GROUP BY a.ledger_id
|
|
25
|
+
),
|
|
26
|
+
reference AS (
|
|
27
|
+
SELECT a.ledger_id, a.cost_micros AS cost
|
|
28
|
+
FROM lane_alternatives a
|
|
29
|
+
JOIN lane_requests r ON r.id = a.ledger_id
|
|
30
|
+
JOIN lane_members m ON m.lane_id = r.lane AND m.model_id = a.model AND m.position = 1
|
|
31
|
+
WHERE a.priced = 1
|
|
32
|
+
),
|
|
33
|
+
selection AS (
|
|
34
|
+
SELECT r.lane AS lane,
|
|
35
|
+
count(*) AS requests,
|
|
36
|
+
sum(r.cost_micros) AS spentMicros,
|
|
37
|
+
sum(d.cost) AS dearestMicros,
|
|
38
|
+
sum(coalesce(ref.cost, 0)) AS referenceMicros,
|
|
39
|
+
count(ref.cost) AS referenceRequests,
|
|
40
|
+
-- From the same scoped set as every other column. A separate
|
|
41
|
+
-- subquery here once read the whole ledger, which for a tenant-scoped
|
|
42
|
+
-- caller counted other tenants' members. Raised in review of PR #63.
|
|
43
|
+
count(DISTINCT r.model) AS membersUsed
|
|
44
|
+
FROM lane_requests r
|
|
45
|
+
-- Only requests whose alternatives were recorded. A lane request from
|
|
46
|
+
-- before 0027 has none, and counting it as "saved nothing" would drag
|
|
47
|
+
-- every percentage toward zero while looking like a measurement.
|
|
48
|
+
JOIN dearest d ON d.ledger_id = r.id
|
|
49
|
+
LEFT JOIN reference ref ON ref.ledger_id = r.id
|
|
50
|
+
GROUP BY r.lane
|
|
51
|
+
),
|
|
52
|
+
-- The cache's own term (#281): a hit reaches no provider and records no
|
|
53
|
+
-- alternatives, so it is outside the selection comparison above and
|
|
54
|
+
-- counted here from the same scoped set, against list.
|
|
55
|
+
cache AS (
|
|
56
|
+
SELECT r.lane AS lane, count(*) AS cacheHits, sum(r.list_micros - r.cost_micros) AS cacheMicros
|
|
57
|
+
FROM lane_requests r WHERE r.cache_status = 'HIT' GROUP BY r.lane
|
|
58
|
+
)
|
|
59
|
+
SELECT lanes.lane AS lane,
|
|
60
|
+
coalesce(s.requests, 0) AS requests,
|
|
61
|
+
coalesce(s.spentMicros, 0) AS spentMicros,
|
|
62
|
+
coalesce(s.dearestMicros, 0) AS dearestMicros,
|
|
63
|
+
coalesce(s.referenceMicros, 0) AS referenceMicros,
|
|
64
|
+
coalesce(s.referenceRequests, 0) AS referenceRequests,
|
|
65
|
+
coalesce(s.membersUsed, 0) AS membersUsed,
|
|
66
|
+
coalesce(c.cacheHits, 0) AS cacheHits,
|
|
67
|
+
coalesce(c.cacheMicros, 0) AS cacheMicros
|
|
68
|
+
FROM (SELECT lane FROM selection UNION SELECT lane FROM cache) lanes
|
|
69
|
+
LEFT JOIN selection s ON s.lane = lanes.lane
|
|
70
|
+
LEFT JOIN cache c ON c.lane = lanes.lane
|
|
71
|
+
ORDER BY lanes.lane`
|
|
72
|
+
);
|
|
73
|
+
const { results } = await statement.bind(month, tenantId ?? null, laneId ?? null).all();
|
|
74
|
+
return results;
|
|
75
|
+
}
|
|
76
|
+
|
|
77
|
+
export {
|
|
78
|
+
laneSavings
|
|
79
|
+
};
|