Skip to main content

[bash-script] vsz, rss logger for one process

#!/bin/bash
BIN_PROC_FILES="my_bin"
LOG_FILE=/tmp/memlog.log
RM_LOG_FILE=0
G_STDOUTPUT=1

function write_log
{
  CUR_TIME_STR=`date +"%Y%m%d_%H%M%S"`
  if [ ${G_STDOUTPUT} -gt 0 ]; then
    echo "${CUR_TIME_STR} $1"
  fi
  if [ "${LOG_FILE}" != "" ] ; then
    echo "${CUR_TIME_STR} $1" >> ${LOG_FILE}
  fi
}

function loop_proc
{
  for CUR_BIN in ${BIN_PROC_FILES}
  do
    PID_SNIPER=`pidof ${CUR_BIN}`
    if [ "${PID_SNIPER}" != "" ]; then
      MEM_STATS=`ps --pid ${PID_SNIPER} -o rss,vsz`
      LINE_CNT=`echo "${MEM_STATS}" | wc -l`
      if [ ${LINE_CNT} -gt 1 ]; then
        MEM_STAT=`echo "${MEM_STATS}" |tail -n 1`
      fi
      write_log "${CUR_BIN} ${PID_SNIPER} ${MEM_STAT}"
    fi
  done
}

if [ ${RM_LOG_FILE} -gt 0 ]; then
  if [ "${LOG_FILE}" != "" ] ; then
    rm ${LOG_FILE}
  fi
fi

while true
do
  loop_proc
  sleep 1
done

Comments

Popular posts from this blog

sanitize on in CMakeLists.txt

1. CMakeLists.txt set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -fsanitize=address") add_compile_options(-fsanitize=address) If a subdir is set with add_subdirectory and a module that does not support sanitize is imported from the path and needs to be removed, add -fno-sanitize=address to add_compile_options in CMakeLists.txt of the subdir to remove it. Basically, if you enable sanitize, sanitize will adjust ulimit's core to 0, so core will not occur. You can do this by putting the following part as an environment variable. (Below is the part you put in systemd's service) 2. service file in systemd [Service] LimitCORE=infinity Environment="ASAN_OPTIONS=handle_segv=0:handle_abort=1:abort_on_error=1:disable_coredump=0"

[bash-script] get_info.sh print cpu allocation infomation of threads

1) way1 ps xH -o 'pid tid cmd comm pcpu psr' 2) way2 #!/bin/bash TMP_FILE=$(mktemp /tmp/abc-script.XXXXXX) echo ${TMP_FILE} ps xH -o 'pid tid cmd comm pcpu ' > ${TMP_FILE} function make_header() {   echo -n "$1"   echo " psr" } function make_values() {   echo -n "$1"   STATUS_FILE=/proc/${2}/task/${3}/status   if [ -f ${STATUS_FILE} ] ; then     CPU_SET=`cat ${STATUS_FILE}  |grep Cpus_allowed_list |awk ' { print $2 } '`   else     CPU_SET="None"   fi   echo " ${CPU_SET}" } let IDX=0 while IFS='' read -r line || [[ -n "$line" ]]; do    LINE_SEP=( $line )    MY_PID=${LINE_SEP[0]}    MY_TID=${LINE_SEP[1]}    if [ ${IDX} -eq 0 ]; then      SZ_HEADER=`make_header "$line"`      echo ${SZ_HEADER}    else      SZ_VALUES=`make_values "$line" ${MY_PID} ${MY_TID}`      echo ${SZ_VALUES}...

e1000e 0000:00:1f.6 eno1: Detected Hardware Unit Hang

 make a file: /usr/local/etc/ethtool.sh (eno1 is your mgmt i/f, vmbr0 is your bridge i/f.) #!/bin/bash ethtool -K eno1 gso off gro on tso off tx on rx on rxvlan on txvlan on sg on && ethtool -K vmbr0 gso off gro on tso off tx on rx on rxvlan on txvlan on sg on ethtool -G eno1 rx 4096 if you want to activate on reboot. /etc/systemd/system/ethtool.service [Unit] Description=ethtool script [Service] WorkingDirectory=/usr/local/etc/ ExecStart=/usr/local/etc/ethtool.sh [Install] WantedBy=multi-user.target then systemctl enable ethtool.service && systemctl start ethtool.service refs: https://forum.proxmox.com/threads/e1000-driver-hang.58284/page-8#post-390709