From a948f1a659e7c7ecadd27bdc66ba25f4a31a4019 Mon Sep 17 00:00:00 2001
From: Nicolas Dupont <npg.dupont@gmail.com>
Date: Thu, 15 Jan 2026 09:54:30 +0100
Subject: [PATCH 1/3] Run dataset publishing sequentially

Large dataset uploads (500MB+) running in parallel caused memory issues
 and silent failures on a machine with 2GB ram
---
 scripts/dataset/publish/index.js | 29 ++++++++++++++++-------------
 1 file changed, 16 insertions(+), 13 deletions(-)

diff --git a/scripts/dataset/publish/index.js b/scripts/dataset/publish/index.js
index ec8906173..367e930ba 100644
--- a/scripts/dataset/publish/index.js
+++ b/scripts/dataset/publish/index.js
@@ -22,26 +22,29 @@ export default async function publishRelease({ archivePath, releaseDate, stats }
     throw new Error('No publishing platform configured. Please configure at least one of: GitHub (OTA_ENGINE_GITHUB_TOKEN), GitLab (OTA_ENGINE_GITLAB_TOKEN), or data.gouv.fr (OTA_ENGINE_DATAGOUV_API_KEY + datasetId or organizationIdOrSlug in config).');
   }
 
-  const results = await Promise.allSettled(platforms.map(async platform => {
-    const url = await platform.publish();
-
-    return { platform: platform.name, url };
-  }));
-
-  const succeeded = results.filter(result => result.status === 'fulfilled');
-  const failed = results.filter(result => result.status === 'rejected');
+  const succeeded = [];
+  const failed = [];
+
+  // Execute publications sequentially to avoid memory issues with large file uploads
+  for (const platform of platforms) {
+    try {
+      const url = await platform.publish();
+
+      succeeded.push({ platform: platform.name, url });
+    } catch (error) {
+      failed.push({ platform: platform.name, error });
+    }
+  }
 
   if (failed.length) {
     let errorMessage = !succeeded.length ? 'All platforms failed to publish:' : 'Some platforms failed to publish:';
 
-    failed.forEach(rejectedResult => {
-      const index = results.indexOf(rejectedResult);
-
-      errorMessage += `\n  - ${platforms[index].name}: ${rejectedResult.reason.message}`;
+    failed.forEach(({ platform, error }) => {
+      errorMessage += `\n  - ${platform}: ${error.message}`;
     });
 
     logger.error(errorMessage);
   }
 
-  return succeeded.map(result => result.value);
+  return succeeded;
 }

From 1b420881e6ebaa3006576e37fdae4add0605052e Mon Sep 17 00:00:00 2001
From: Nicolas Dupont <npg.dupont@gmail.com>
Date: Thu, 15 Jan 2026 10:16:21 +0100
Subject: [PATCH 2/3] Add changelog entry

---
 CHANGELOG.md | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index a072d3337..caa5dfcba 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,6 +2,14 @@
 
 All changes that impact users of this module are documented in this file, in the [Common Changelog](https://common-changelog.org) format with some additional specifications defined in the CONTRIBUTING file. This codebase adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## Unreleased [patch]
+
+> Development of this release was supported by [Reset Tech](https://www.reset.tech).
+
+### Fixed
+
+- Fix dataset publishing failure for large files by executing uploads sequentially instead of in parallel
+
 ## 10.3.2 - 2026-01-14
 
 > Development of this release was supported by [Reset Tech](https://www.reset.tech).

From d2c86cf9f265bd868c52eb623d6a7db553b14f94 Mon Sep 17 00:00:00 2001
From: Nicolas Dupont <npg.dupont@gmail.com>
Date: Thu, 15 Jan 2026 16:58:06 +0100
Subject: [PATCH 3/3] Stream dataset file for GitHub release upload

Avoid loading the entire archive into memory by using createReadStream instead of readFileSync. This prevents out-of-memory crashes when publishing large datasets on memory-constrained machines.
---
 scripts/dataset/publish/github/index.js | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/dataset/publish/github/index.js b/scripts/dataset/publish/github/index.js
index 5aaec8ad5..b6696cd12 100644
--- a/scripts/dataset/publish/github/index.js
+++ b/scripts/dataset/publish/github/index.js
@@ -30,7 +30,7 @@ export default async function publish({ archivePath, releaseDate, stats }) {
   logger.info('Uploading release asset…');
 
   await octokit.rest.repos.uploadReleaseAsset({
-    data: fsApi.readFileSync(archivePath),
+    data: fsApi.createReadStream(archivePath),
     headers: {
       'content-type': 'application/zip',
       'content-length': fsApi.statSync(archivePath).size,