Intermittent unknown errors, codes 134 and 137, and JS Out of Memory on Worker Build
Workers
Hello, I've been running into failed builds consistently with a Nuxt app. The build works fine locally. What's odd is for the OOM error, it's showing only 2GB memory even though the builder should have 8GB (paid plan). The build works fine locally using
export NODE_OPTIONS='--max-old-space-size=8192'
export NODE_OPTIONS='--max-old-space-size=8192'
Job ID 38d1f420-1e77-4ffb-8e11-e7f4db8f85d9 2026-01-21T16:48:52.931Z An internal error occurred. Please retry your build. If this problem persists, contact Cloudflare support.
Job ID 58179326-043e-49dd-91f8-6ef4de6ce3de 2026-01-21T16:57:12.443Z Killed 2026-01-21T16:57:12.513Z ELIFECYCLE Command failed with exit code 137. 2026-01-21T16:57:12.600Z Failed: error occurred while running build command
Job ID 9b67d6c1-3ca4-4dbf-899f-cc61f9aa45b3 00:27:50.827 <--- Last few GCs ---> 00:27:50.827 00:27:50.827 [1461:0xfe41000] 69948 ms: Mark-Compact 1997.9 (2080.3) -> 1991.8 (2086.3) MB, pooled: 0 MB, 1282.00 / 0.00 ms (average mu = 0.586, current mu = 0.219) allocation failure; scavenge might not succeed 00:27:50.827 [1461:0xfe41000] 71239 ms: Mark-Compact 2008.2 (2087.0) -> 2006.3 (2098.8) MB, pooled: 0 MB, 1256.79 / 0.00 ms (average mu = 0.407, current mu = 0.027) allocation failure; scavenge might not succeed 00:27:50.827 00:27:50.827 <--- JS stacktrace ---> 00:27:50.828 00:27:50.828 FATAL ERROR: Reached heap limit Allocation failed - JavaScript heap out of memory 00:27:50.828 ----- Native stack trace ----- 00:27:50.828 00:27:50.831 1: 0xe13fde node::OOMErrorHandler(char const, v8::OOMDetails const&) [node] 00:27:50.832 2: 0x11d5070 v8::Utils::ReportOOMFailure(v8::internal::Isolate, char const*, v8::OOMDetails const&) [node]