perf(npm): load npm resolution snapshot directly from lockfile #28647

nathanwhit · 2025-03-27T16:28:11Z

Currently the new lockfile version is gated behind an unstable flag (--unstable-lockfile-v5) until the next minor release, where it will become the default.

The main motivation here is that it improves startup performance when using the global cache or --node-modules-dir=auto.

In a create-next-app project, running an empty file:

❯ hyperfine --warmup 25 -N --setup "rm -f deno.lock" "deno run --node-modules-dir=auto -A empty.js" "deno-this-pr run --node-modules-dir=auto -A empty.js" "deno-this-pr run --node-modules-dir=auto --unstable-lockfile-v5 empty.js" "deno run --node-modules-dir=manual -A empty.js" "deno-this-pr run --node-modules-dir=manual -A empty.js"
Benchmark 1: deno run --node-modules-dir=auto -A empty.js
  Time (mean ± σ):     247.6 ms ±   1.7 ms    [User: 228.7 ms, System: 19.0 ms]
  Range (min … max):   245.5 ms … 251.5 ms    12 runs

Benchmark 2: deno-this-pr run --node-modules-dir=auto -A empty.js
  Time (mean ± σ):     169.8 ms ±   1.0 ms    [User: 152.9 ms, System: 17.9 ms]
  Range (min … max):   168.9 ms … 172.5 ms    17 runs

Benchmark 3: deno-this-pr run --node-modules-dir=auto --unstable-lockfile-v5 empty.js
  Time (mean ± σ):      16.2 ms ±   0.7 ms    [User: 12.3 ms, System: 5.7 ms]
  Range (min … max):    15.2 ms …  19.2 ms    185 runs

Benchmark 4: deno run --node-modules-dir=manual -A empty.js
  Time (mean ± σ):      16.2 ms ±   0.8 ms    [User: 11.6 ms, System: 5.5 ms]
  Range (min … max):    14.9 ms …  19.7 ms    187 runs

Benchmark 5: deno-this-pr run --node-modules-dir=manual -A empty.js
  Time (mean ± σ):      16.0 ms ±   0.9 ms    [User: 12.0 ms, System: 5.5 ms]
  Range (min … max):    14.8 ms …  22.3 ms    190 runs

  Warning: Statistical outliers were detected. Consider re-running this benchmark on a quiet system without any interferences from other programs. It might help to use the '--warmup' or '--prepare' options.

Summary
  deno-this-pr run --node-modules-dir=manual -A empty.js ran
    1.01 ± 0.08 times faster than deno run --node-modules-dir=manual -A empty.js
    1.01 ± 0.07 times faster than deno-this-pr run --node-modules-dir=auto --unstable-lockfile-v5 empty.js
   10.64 ± 0.60 times faster than deno-this-pr run --node-modules-dir=auto -A empty.js
   15.51 ± 0.88 times faster than deno run --node-modules-dir=auto -A empty.js

When using the new lockfile version, this leads to a 15.5x faster startup time compared to the current deno version.

Install times benefit as well, though to a lesser degree.

deno install on a create-next-app project, with everything cached (just setting up node_modules from scratch):

❯ hyperfine --warmup 5 -N --prepare "rm -rf node_modules" --setup "rm -rf deno.lock" "deno i" "deno-this-pr i" "deno-this-pr i --unstable-lockfile-v5"
Benchmark 1: deno i
  Time (mean ± σ):     464.4 ms ±   8.8 ms    [User: 227.7 ms, System: 217.3 ms]
  Range (min … max):   452.6 ms … 478.3 ms    10 runs

Benchmark 2: deno-this-pr i
  Time (mean ± σ):     368.8 ms ±  22.0 ms    [User: 150.8 ms, System: 198.1 ms]
  Range (min … max):   344.8 ms … 397.6 ms    10 runs

Benchmark 3: deno-this-pr i --unstable-lockfile-v5
  Time (mean ± σ):     211.9 ms ±  17.1 ms    [User: 7.1 ms, System: 177.2 ms]
  Range (min … max):   191.3 ms … 233.4 ms    10 runs

Summary
  deno-this-pr i --unstable-lockfile-v5 ran
    1.74 ± 0.17 times faster than deno-this-pr i
    2.19 ± 0.18 times faster than deno i

With lockfile v5, a 2.19x faster install time compared to the current deno.

Copilot

Copilot reviewed 71 out of 71 changed files in this pull request and generated no comments.

Comments suppressed due to low confidence (3)

cli/npm/installer/common/bin_entries.rs:18

[nitpick] Changing from a borrowed &str key to a String key may incur additional allocation overhead; please verify that this impact is acceptable from a performance standpoint.

seen_names: HashMap<String, &'a NpmPackageId>,

cli/args/lockfile.rs:36

[nitpick] Consider adding inline documentation for the 'use_lockfile_v5' field to clearly describe when and why the lockfile v5 behavior is enabled.

pub use_lockfile_v5: bool,

cli/main.rs:187

[nitpick] Verify that converting the npm registry info provider using as_npm_registry_api() produces the correct API expected by the LSP start function.

lsp::start(Arc::new(factory.npm_registry_info_provider()?.as_npm_registry_api())).await

tests/integration/npm_tests.rs

marvinhagemeister

Code LGTM to me, but I'll let @dsherret do the final review as he's way more familiar with these areas in Deno.

cli/args/flags.rs

cli/args/lockfile.rs

nathanwhit

self review

nathanwhit · 2025-04-04T18:32:19Z

cli/args/mod.rs

@@ -1159,9 +1199,10 @@ impl CliOptions {
          "fmt-component",
          "fmt-sql",
          "lazy-dynamic-imports",
-          "lazy-npm-caching",
+          "npm-lazy-caching",


drive by fix

cli/factory.rs

nathanwhit · 2025-04-04T18:47:39Z

cli/factory.rs

-    Ok(self.workspace_factory()?.npmrc()?)
-  }
-
-  pub fn resolver_factory(&self) -> Result<&Arc<CliResolverFactory>, AnyError> {


not sure why i moved these

nathanwhit · 2025-04-04T18:49:28Z

cli/lsp/tsc.rs

+  struct DefaultRegistry;
+
+  #[async_trait::async_trait(?Send)]
+  impl deno_npm::registry::NpmRegistryApi for DefaultRegistry {
+    async fn package_info(
+      &self,
+      _name: &str,
+    ) -> Result<Arc<NpmPackageInfo>, NpmRegistryPackageInfoLoadError> {
+      Ok(Arc::new(NpmPackageInfo::default()))
+    }
+  }
+
+  fn default_registry() -> Arc<dyn NpmRegistryApi + Send + Sync> {
+    Arc::new(DefaultRegistry)
+  }
+


this is duplicated across like 4 files in lsp tests, should find somewhere shared to put it

cli/npm/installer/global.rs

tests/specs/install/future_install_node_modules/__test__.jsonc

tests/specs/lockfile/adding_npm_dep/lock02.out

tests/specs/task/signals/__test__.jsonc

tests/testdata/npm/peer_deps_with_copied_folders/main.out

tests/util/server/src/builders.rs

dsherret

LGTM! This is going to be great.

Fixes #27264. Fixes #28161. Currently the new lockfile version is gated behind an unstable flag (`--unstable-lockfile-v5`) until the next minor release, where it will become the default. The main motivation here is that it improves startup performance when using the global cache or `--node-modules-dir=auto`. In a create-next-app project, running an empty file: ``` ❯ hyperfine --warmup 25 -N --setup "rm -f deno.lock" "deno run --node-modules-dir=auto -A empty.js" "deno-this-pr run --node-modules-dir=auto -A empty.js" "deno-this-pr run --node-modules-dir=auto --unstable-lockfile-v5 empty.js" "deno run --node-modules-dir=manual -A empty.js" "deno-this-pr run --node-modules-dir=manual -A empty.js" Benchmark 1: deno run --node-modules-dir=auto -A empty.js Time (mean ± σ): 247.6 ms ± 1.7 ms [User: 228.7 ms, System: 19.0 ms] Range (min … max): 245.5 ms … 251.5 ms 12 runs Benchmark 2: deno-this-pr run --node-modules-dir=auto -A empty.js Time (mean ± σ): 169.8 ms ± 1.0 ms [User: 152.9 ms, System: 17.9 ms] Range (min … max): 168.9 ms … 172.5 ms 17 runs Benchmark 3: deno-this-pr run --node-modules-dir=auto --unstable-lockfile-v5 empty.js Time (mean ± σ): 16.2 ms ± 0.7 ms [User: 12.3 ms, System: 5.7 ms] Range (min … max): 15.2 ms … 19.2 ms 185 runs Benchmark 4: deno run --node-modules-dir=manual -A empty.js Time (mean ± σ): 16.2 ms ± 0.8 ms [User: 11.6 ms, System: 5.5 ms] Range (min … max): 14.9 ms … 19.7 ms 187 runs Benchmark 5: deno-this-pr run --node-modules-dir=manual -A empty.js Time (mean ± σ): 16.0 ms ± 0.9 ms [User: 12.0 ms, System: 5.5 ms] Range (min … max): 14.8 ms … 22.3 ms 190 runs Warning: Statistical outliers were detected. Consider re-running this benchmark on a quiet system without any interferences from other programs. It might help to use the '--warmup' or '--prepare' options. Summary deno-this-pr run --node-modules-dir=manual -A empty.js ran 1.01 ± 0.08 times faster than deno run --node-modules-dir=manual -A empty.js 1.01 ± 0.07 times faster than deno-this-pr run --node-modules-dir=auto --unstable-lockfile-v5 empty.js 10.64 ± 0.60 times faster than deno-this-pr run --node-modules-dir=auto -A empty.js 15.51 ± 0.88 times faster than deno run --node-modules-dir=auto -A empty.js ``` When using the new lockfile version, this leads to a 15.5x faster startup time compared to the current deno version. Install times benefit as well, though to a lesser degree. `deno install` on a create-next-app project, with everything cached (just setting up node_modules from scratch): ``` ❯ hyperfine --warmup 5 -N --prepare "rm -rf node_modules" --setup "rm -rf deno.lock" "deno i" "deno-this-pr i" "deno-this-pr i --unstable-lockfile-v5" Benchmark 1: deno i Time (mean ± σ): 464.4 ms ± 8.8 ms [User: 227.7 ms, System: 217.3 ms] Range (min … max): 452.6 ms … 478.3 ms 10 runs Benchmark 2: deno-this-pr i Time (mean ± σ): 368.8 ms ± 22.0 ms [User: 150.8 ms, System: 198.1 ms] Range (min … max): 344.8 ms … 397.6 ms 10 runs Benchmark 3: deno-this-pr i --unstable-lockfile-v5 Time (mean ± σ): 211.9 ms ± 17.1 ms [User: 7.1 ms, System: 177.2 ms] Range (min … max): 191.3 ms … 233.4 ms 10 runs Summary deno-this-pr i --unstable-lockfile-v5 ran 1.74 ± 0.17 times faster than deno-this-pr i 2.19 ± 0.18 times faster than deno i ``` With lockfile v5, a 2.19x faster install time compared to the current deno.

nathanwhit added the ci-draft Run the CI on draft PRs. label Mar 27, 2025

nathanwhit force-pushed the packument-lockfile-v2 branch 3 times, most recently from 1a58769 to 8cd172b Compare April 1, 2025 19:11

nathanwhit added 16 commits April 1, 2025 12:22

well, it compiles

9c2f14d

avoid making everything async

183587a

it works

2dda5ba

fix, start fixing tests

583846c

update more tests

770ed94

fmt, cleanup, point to git

03b2f82

rebase fixup

e113f16

update tests

17deed1

appease clippy

52cb9b5

fmt

109defa

appease clippy

2dc33c9

fix lockfile

2922999

update for changes

8009c2c

update tests pt 1

3162c20

fix lockfile

cc0a74d

wire up unstable flag

2d42c85

nathanwhit force-pushed the packument-lockfile-v2 branch from 8cd172b to 2d42c85 Compare April 1, 2025 21:35

nathanwhit added 4 commits April 1, 2025 15:12

add flag

94b8ec8

fix

4f9b7da

finish updating tests

98828a4

Merge main into packument-lockfile-v2

fbf9f95

nathanwhit requested a review from Copilot April 3, 2025 21:47

use published versions

852f2b0

Copilot AI reviewed Apr 3, 2025

View reviewed changes

nathanwhit added 2 commits April 3, 2025 14:55

omit tarball url from lockfile if default

d9e5c2f

Merge upstream/main into packument-lockfile-v2

3af7f22

marvinhagemeister reviewed Apr 4, 2025

View reviewed changes

tests/integration/npm_tests.rs Outdated Show resolved Hide resolved

marvinhagemeister reviewed Apr 4, 2025

View reviewed changes

nathanwhit added 2 commits April 4, 2025 11:23

Merge upstream/main into packument-lockfile-v2

953c061

Merge upstream/main into packument-lockfile-v2

bf9de28

nathanwhit requested a review from dsherret April 4, 2025 18:29

nathanwhit commented Apr 4, 2025

View reviewed changes

cli/args/flags.rs Outdated Show resolved Hide resolved

nathanwhit commented Apr 4, 2025

View reviewed changes

cli/args/lockfile.rs Outdated Show resolved Hide resolved

nathanwhit commented Apr 4, 2025

View reviewed changes

cli/args/lockfile.rs Outdated Show resolved Hide resolved

nathanwhit commented Apr 4, 2025

View reviewed changes

nathanwhit added 15 commits April 7, 2025 09:36

Merge main into packument-lockfile-v2

2c029bb

don't get extra info in global installer

f1ea8aa

cleanup

b039700

cleanup

ec4c189

remove unused test files

9a5f98b

restore test

6ebbb98

reduce duplication

a90c7e5

appease clippy

c0683e6

fix

f850f5c

fix, better error message, add tests

7a9d5a4

remove unused output file

6a5c06a

refactor cli options, remove tokio runtime hack

2951120

cleanup + fix test

1225613

remove commented out code

80b0d0f

revert unintentional change

6af40ea

dsherret approved these changes Apr 8, 2025

View reviewed changes

nathanwhit enabled auto-merge (squash) April 8, 2025 01:50

nathanwhit merged commit 780b741 into denoland:main Apr 8, 2025
17 checks passed

nathanwhit deleted the packument-lockfile-v2 branch April 9, 2025 18:39

dsherret mentioned this pull request Apr 15, 2025

Make lockfile v5 the default #28916

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

perf(npm): load npm resolution snapshot directly from lockfile #28647

perf(npm): load npm resolution snapshot directly from lockfile #28647

Uh oh!

nathanwhit commented Mar 27, 2025 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

marvinhagemeister left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

nathanwhit left a comment

Uh oh!

nathanwhit Apr 4, 2025

Uh oh!

Uh oh!

nathanwhit Apr 4, 2025

Uh oh!

nathanwhit Apr 4, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

dsherret left a comment

Uh oh!

Uh oh!

Uh oh!

perf(npm): load npm resolution snapshot directly from lockfile #28647

perf(npm): load npm resolution snapshot directly from lockfile #28647

Uh oh!

Conversation

nathanwhit commented Mar 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

marvinhagemeister left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

nathanwhit left a comment

Choose a reason for hiding this comment

Uh oh!

nathanwhit Apr 4, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

nathanwhit Apr 4, 2025

Choose a reason for hiding this comment

Uh oh!

nathanwhit Apr 4, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

dsherret left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

nathanwhit commented Mar 27, 2025 •

edited

Loading