From: Peter Amstutz <peter.amstutz@curii.com>
Date: Tue, 19 Apr 2022 19:40:56 +0000 (-0400)
Subject: 17301: Report warning about OOM killer when exit code 137
X-Git-Tag: 2.5.0~196^2~6
X-Git-Url: https://git.arvados.org/arvados.git/commitdiff_plain/6358388ad9f7676aa3b1ab149c9fbde4716929e5

17301: Report warning about OOM killer when exit code 137

Arvados-DCO-1.1-Signed-off-by: Peter Amstutz <peter.amstutz@curii.com>
---

diff --git a/sdk/cwl/arvados_cwl/arvcontainer.py b/sdk/cwl/arvados_cwl/arvcontainer.py
index e2c2f2e67b..c85443a23a 100644
--- a/sdk/cwl/arvados_cwl/arvcontainer.py
+++ b/sdk/cwl/arvados_cwl/arvcontainer.py
@@ -392,6 +392,10 @@ class ArvadosContainer(JobBase):
                     processStatus = "success"
                 else:
                     processStatus = "permanentFail"
+
+                if rcode == 137:
+                    logger.warning("%s job was killed on the compute instance.  The most common reason is that it attempted to allocate too much RAM and was targeted by the Out Of Memory (OOM) killer.  Try resubmitting with a higher 'ramMin'.",
+                                 self.arvrunner.label(self))
             else:
                 processStatus = "permanentFail"