diff --git a/oneflow/python/deprecated/init_cluster_env.py b/oneflow/python/deprecated/init_cluster_env.py index 241806d894988a4b8d637b0eb66437d154071335..e37f278697d670cc6c7eb10af94b7c04e63d7a15 100644 --- a/oneflow/python/deprecated/init_cluster_env.py +++ b/oneflow/python/deprecated/init_cluster_env.py @@ -140,7 +140,16 @@ def _SendBinaryAndConfig2Worker( + ' 1>/dev/null 2>&1 </dev/null & "' ) _SystemCall(ssh_prefix + oneflow_cmd) - _SystemCall(ssh_prefix + "ps aux") + proc = subprocess.Popen( + ssh_prefix + "ps aux", + stdout=subprocess.PIPE, + stderr=subprocess.PIPE, + encoding="utf-8", + shell=True, + ) + outs, errs = proc.communicate(timeout=5) + print(outs) + assert "oneflow_worker" in str(outs), "fail to start oneflow_worker" print("oneflow worker initialized:", machine.addr, flush=True)