Commit 3441f235 authored by unknown's avatar unknown
Browse files

BUG#13987 Cluster: Loss of data nodes can cause high CPU usage from ndb_mgmd 

A further reduction in CPU usage by ndb_mgmd with failed data nodes.


ndb/src/common/util/socket_io.cpp:
  Use a nonblocking socket to peek at the buffer to know how much to read for socket_readln.
  
  This saves us a lot of syscalls. By a lot, I've observed about halving the CPU
  usage of ndb_mgmd with 1 failed node on a 6 node cluster.
parent e42692c0

ndb/src/common/util/socket_io.cpp

+39 −31
Original line number Diff line number Diff line
@@ -53,6 +53,10 @@ readln_socket(NDB_SOCKET_TYPE socket, int timeout_millis, 
  if(buflen <= 1)

    return 0;



  int sock_flags= fcntl(socket, F_GETFL);

  if(fcntl(socket, F_SETFL, sock_flags | O_NONBLOCK) == -1)

    return -1;



  fd_set readset;

  FD_ZERO(&readset);

  FD_SET(socket, &readset);
@@ -62,44 +66,48 @@ readln_socket(NDB_SOCKET_TYPE socket, int timeout_millis, 
  timeout.tv_usec = (timeout_millis % 1000) * 1000;



  const int selectRes = select(socket + 1, &readset, 0, 0, &timeout);

  if(selectRes == 0)

  if(selectRes == 0){

    return 0;

  }



  if(selectRes == -1){

    fcntl(socket, F_SETFL, sock_flags);

    return -1;

  }



  int pos = 0; buf[pos] = 0;

  while(true){

    const int t = recv(socket, &buf[pos], 1, 0);

    if(t != 1){

  buf[0] = 0;

  const int t = recv(socket, buf, buflen, MSG_PEEK);



  if(t < 1)

  {

    fcntl(socket, F_SETFL, sock_flags);

    return -1;

  }

    if(buf[pos] == '\n'){

      buf[pos] = 0;



      if(pos > 0 && buf[pos-1] == '\r'){

	pos--;

	buf[pos] = 0;

  for(int i=0; i< t;i++)

  {

    if(buf[i] == '\n'){

      recv(socket, buf, i+1, 0);

      buf[i] = 0;



      if(i > 0 && buf[i-1] == '\r'){

        i--;

        buf[i] = 0;

      }



      return pos;

      fcntl(socket, F_SETFL, sock_flags);

      return t;

    }

    pos++;

    if(pos == (buflen - 1)){

      buf[pos] = 0;

      return buflen;

  }



    FD_ZERO(&readset);

    FD_SET(socket, &readset);

    timeout.tv_sec  = (timeout_millis / 1000);

    timeout.tv_usec = (timeout_millis % 1000) * 1000;

    const int selectRes = select(socket + 1, &readset, 0, 0, &timeout);

    if(selectRes != 1){

      return -1;

    }

  if(t == (buflen - 1)){

    recv(socket, buf, t, 0);

    buf[t] = 0;

    fcntl(socket, F_SETFL, sock_flags);

    return buflen;

  }



  return 0;

}



extern "C"