iometer 1.1.0 版本包含了3种Write IO Data Pattern。分别是：Repeating bytes、Pseudo-Random (2006.07.27及其之前的版本默认) 和 Full random。实际上这三种IO数据类型的代码在1.1.0 rc1 就已经合入。

-------------------

Version 1.1.0 (RC1)
===================
- New features:
- (All) Different data randomization options are now available via a drop down menu in the GUI. Options are: "Repeating bytes","Pseudo-Random" (used by version 2006.07.27 and older) and "Full random".
- (All) Randomization seeding support (by Ryan Bever) was added to define a fixed seed for the random number generator. A fixed seed can be specified on a per Manager or per Worker basis in the Disk Targets page of the GUI.
..........（省略多字，详细内容将查阅：http://sourceforge.net/p/iometer/svn/HEAD/tree/trunk/IOmeter/CHANGELOG.txt）

-------------------

IOTargetDisk.cpp 文件：

//
// Logical drives are accessed through a file on the drive.  This will create
// the file by writing in units of "bytes" beginning at "*prepare_offset".  It
// maintains a simple I/O queue of depth PREPARE_QDEPTH, which it keeps as
// full as possible.  It keeps queueing up new I/Os until the disk fills up,
// or the user-specified file size (starting_sector + maximum_size) is reached,
// or a STOP_PREPARE message is received (reflected in "*test_state").  It
// returns only when all queued I/Os have completed. 
// Return value is TRUE for success, FALSE if any error occurred.
//
BOOL TargetDisk::Prepare(DWORDLONG * prepare_offset, volatile TestState * test_state, int sector_size, unsigned char* _random_data_buffer, long long _random_datat_buffer_size)
{
 BOOL write_ok;
 int num_outstanding;
 DWORD bytes_written;
 OVERLAPPED olap[PREPARE_QDEPTH];
 BOOL busy[PREPARE_QDEPTH];
 BOOL retval;
 int i;
 void *buffer = NULL;
 DWORD bytes; // static to aid switch to single sectors to finish up preparing
 BOOL onSingleSectors = FALSE; // mark if a switch needs to be made to single sector writes
 BOOL singleSectorSwitch = FALSE; // mark that single sectors are being used to write
 DWORDLONG rand_seed;
#ifdef IOMTR_SETTING_LINUX_LIBAIO
  struct statfs fsInfo;
  int statResult;
  int fd = -1;
  fd = ((struct File *)disk_file)->fd;
#endif

 // Save current spec.random so that it can be reset back to the same seed value after preparing disks
 // This allows the PRNG to be the same when doing the actual IO, whether the disk is prepped or not and the user specifies fixed seed
 rand_seed = spec.random;

 // Allocate a large (64k for 512 byte sector size) buffer for the preparation.
 bytes = sector_size * 128;
#if defined(IOMTR_OSFAMILY_NETWARE)
 NXMemFree(buffer);
 errno = 0;
 if (!(buffer = NXMemAlloc(bytes, 1)))
#elif defined(IOMTR_OSFAMILY_UNIX)
 free(buffer);
 errno = 0;
 if (!(buffer = valloc(bytes)))
#elif defined(IOMTR_OSFAMILY_WINDOWS)
 VirtualFree(buffer, 0, MEM_RELEASE);
 if (!(buffer = VirtualAlloc(NULL, bytes, MEM_COMMIT, PAGE_READWRITE)))
#else
#warning ===> WARNING: You have to do some coding here to get the port done!
#endif
 {
  cout << "*** Could not allocate buffer to prepare disk." << endl;
  return FALSE;
 }

 switch (spec.DataPattern) {
  case DATA_PATTERN_REPEATING_BYTES:
   // Do nothing here...a new random byte will be chosen below for each IO
   break;
  case DATA_PATTERN_PSEUDO_RANDOM:
   for( DWORD x = 0; x < bytes; x++)
    ((unsigned char*)buffer)[x] = (unsigned char)Rand(0xff);
   break;
  case DATA_PATTERN_FULL_RANDOM:
   //Nothing to do here
   break;
 }

..................

#ifdef IOMTR_SETTING_LINUX_LIBAIO
    statResult = fstatfs(fd, &fsInfo);
    if (statResult < 0) {
     cerr << __FUNCTION__ << ": Couldn‘t statfs logical disk file!" << endl;
    }
    if (fsInfo.f_bfree <= 0 ) {
     write_ok = FALSE;
     break;
    }
#endif

    // If we are still writing and the slot is not busy, start an I/O for
    // this slot.
    if ((*test_state == TestPreparing) && write_ok && !busy[i]) {
     // Set its address.
     olap[i].Offset = (DWORD) * prepare_offset;
     olap[i].OffsetHigh = (DWORD) (*prepare_offset >> 32);

     // Fill the buffer with some new random data so we aren‘t writing all zeros each time
     switch (spec.DataPattern) {
      case DATA_PATTERN_REPEATING_BYTES:
       memset(buffer, rand(), bytes);
       break;
      case DATA_PATTERN_PSEUDO_RANDOM:
       break; // Nothing to do here, buffer was set above
      case DATA_PATTERN_FULL_RANDOM:
       //Buffer offset must be DWORD-aligned in memory, otherwise the transfer fails
       //Choose a pointer into the buffer
       long long mem_offset = (long long)Rand(_random_datat_buffer_size-bytes);

       //See how far it is from being DWORD-aligned
       long long remainder = mem_offset & (sizeof(DWORD) - 1);

       //Align the pointer using the remainder
       mem_offset = mem_offset - remainder;

       buffer = &_random_data_buffer[mem_offset];
       break;
     }

     // Do the asynchronous write.
     if (WriteFile(disk_file, (char *)buffer, bytes, &bytes_written, &(olap[i]))) {
      // It succeeded immediately!
#ifdef _DEBUG
      cout << "Wrote (immediately) " << bytes_written << " of "
          << bytes << " bytes to disk " << spec.name << "." << endl;
#endif
      // Advance the file pointer, but do not mark the slot as busy.
      *prepare_offset += bytes;
     } else {
      // It did not succeed immediately... did it start OK?
      if (GetLastError() == ERROR_IO_PENDING) {
       // It started OK.
#if _DETAILS
       cout << "I/O started successfully for slot #" << i
           << " for " << bytes << " bytes at address "
           << *prepare_offset << endl;
#endif

---------------

Repeating bytes：

Pseudo-Random：

Full random：

• 应用
  

    目前我知道的一个领域情况是，iometer 选择不同IO Data Pattern，可以用于数据压缩算法产品的测试结果的效果宣传上。存储领域，我想近几年的数据重删应用，可能偶尔会用到这些。对比三种方式，可以看出，采用Repeating bytes 方式，是最利于产品宣传的。而iometer 2006.07.27 版本默认是伪随机方式，故，最好采用iometer 1.1.0 GA 版本，或者基于这个版本的增强版本（笔者博客目前有一个20150123A1版本）。

本文出自 “Gedankenwesen” 博客，请务必保留此出处http://yungchin.blog.51cto.com/9614170/1609825

iometer 1.1.0 重要特性说明之Write IO Data Pattern