任务运行失败,Check failed: len > 0
Created by: lyp2github
平台:paddlecloud jobid= job-e6c5bd6b88eb7964 历史类似问题说是因为超时被kill, 但任务明显没有超时 作业提交时间:2018-10-29 15:37:03 作业开始时间:2018-10-29 15:37:40 作业结束时间:2018-10-29 17:12:46 已经执行的时间:01 hour 35 min 01 sec
报错信息: F1029 17:07:19.030912 20059 SocketChannel.cpp:101] Check failed: len > 0 peer=10.75.68.23 curIov=23 iovCnt=117 iovs[curIov].base=0x7f5a0ec8a99d iovs[curIov].iov_len=34467 *** Check failure stack trace: *** @ 0x7f5a43a8af5d google::LogMessage::Fail() @ 0x7f5a43a8ea0c google::LogMessage::SendToLog() @ 0x7f5a43a8aa83 google::LogMessage::Flush() @ 0x7f5a43a8ff1e google::LogMessageFatal::~LogMessageFatal() @ 0x7f5a43866bd1 paddle::readwritev<>() @ 0x7f5a438677e4 paddle::MsgReader::readBlocks() @ 0x7f5a4566459a paddle::ParameterClient2::sendParallel() @ 0x7f5a439d43cc _ZNSt6thread5_ImplISt12_Bind_simpleIFZN6paddle14SyncThreadPool5startEvEUliE_mEEE6_M_runEv @ 0x7f5a9793e8a0 execute_native_thread_routine @ 0x7f5aa88a21c3 start_thread @ 0x7f5aa7eca12d __clone