diff --git a/src/common/di.rs b/src/common/di.rs
index 97f1c790..2e4c8db8 100644
--- a/src/common/di.rs
+++ b/src/common/di.rs
@@ -363,9 +363,9 @@ impl AppServiceFactory {
             if self.config.features.enable_video_thumbnails
                 && FfmpegVideoFrameService::is_available(&ffmpeg_path)
             {
-                let cpus = std::thread::available_parallelism()
-                    .map(|n| n.get())
-                    .unwrap_or(4);
+                // effective_parallelism respects the CFS quota (--cpus), not
+                // just affinity — so ffmpeg fan-out matches the real core budget.
+                let cpus = crate::common::runtime::effective_parallelism();
                 let concurrency = std::env::var("OXICLOUD_VIDEO_THUMBNAIL_CONCURRENCY")
                     .ok()
                     .and_then(|v| v.parse::<usize>().ok())
diff --git a/src/infrastructure/services/image_transcode_service.rs b/src/infrastructure/services/image_transcode_service.rs
index 4501f746..6b0b8bbe 100644
--- a/src/infrastructure/services/image_transcode_service.rs
+++ b/src/infrastructure/services/image_transcode_service.rs
@@ -31,13 +31,14 @@ pub const MAX_TRANSCODE_SIZE: u64 = 5 * 1024 * 1024;
 /// Minimum number of threads in the dedicated transcoding pool
 const MIN_TRANSCODE_THREADS: usize = 2;
 
-/// Compute the number of transcoding threads: half the available CPUs,
-/// with a floor of `MIN_TRANSCODE_THREADS`.  `available_parallelism()`
-/// respects cgroup limits (Docker/K8s) and CPU affinity masks.
+/// Compute the number of transcoding threads: half the available CPUs, with a
+/// floor of `MIN_TRANSCODE_THREADS`. Sized by
+/// [`effective_parallelism`](crate::common::runtime::effective_parallelism),
+/// which respects CPU affinity **and** the CFS quota (Docker/K8s `--cpus`) —
+/// unlike bare `available_parallelism()`, which ignores the quota and would
+/// over-size this CPU-bound pool under a container limit.
 fn transcode_thread_count() -> usize {
-    let cpus = std::thread::available_parallelism()
-        .map(|n| n.get())
-        .unwrap_or(MIN_TRANSCODE_THREADS);
+    let cpus = crate::common::runtime::effective_parallelism();
     (cpus / 2).max(MIN_TRANSCODE_THREADS)
 }
 
diff --git a/src/infrastructure/services/thumbnail_service.rs b/src/infrastructure/services/thumbnail_service.rs
index cbcd2b2a..8d8e37cf 100644
--- a/src/infrastructure/services/thumbnail_service.rs
+++ b/src/infrastructure/services/thumbnail_service.rs
@@ -107,8 +107,11 @@ const STREAM_TO_TEMP_TIMEOUT: Duration = Duration::from_secs(120);
 /// concurrency was halved to keep peak RAM in check. Decodes are now DCT-shrunk
 /// to the thumbnail size (~18–25 MB regardless of source resolution), so the RAM
 /// ceiling no longer forces throttling and we can saturate every core. Override
-/// with `OXICLOUD_THUMBNAIL_DECODE_CONCURRENCY`. `available_parallelism()`
-/// respects cgroup limits (Docker/K8s) and CPU affinity masks.
+/// with `OXICLOUD_THUMBNAIL_DECODE_CONCURRENCY`. Sized by
+/// [`effective_parallelism`](crate::common::runtime::effective_parallelism),
+/// which respects CPU affinity **and** the CFS quota (`--cpus`) — so under a
+/// container quota we don't over-permit concurrent CPU-heavy decodes onto cores
+/// the scheduler can't give us.
 fn max_concurrent_decodes() -> usize {
     if let Some(n) = std::env::var(DECODE_CONCURRENCY_ENV)
         .ok()
@@ -117,10 +120,7 @@ fn max_concurrent_decodes() -> usize {
     {
         return n;
     }
-    let cpus = std::thread::available_parallelism()
-        .map(|n| n.get())
-        .unwrap_or(4);
-    cpus.max(2)
+    crate::common::runtime::effective_parallelism().max(2)
 }
 
 /// Thumbnail service for generating and caching image thumbnails