<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Transitional//EN">
<HTML>
<HEAD>
  <!--
    *************************************************************************************
    * Copyright (c) 2005 VERITAS Software Corporation. All rights reserved. *

      For third-party license citations, please see the bottom of this file.
    *************************************************************************************
  -->
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<meta name="TEMPLATEBASE" content="vx_vxhelp_2005c">
<meta name="LASTUPDATED" content="06/13/05 10:02:06">
<meta name="PRODUCTNAME" content="VERITAS Software Documentation">
<meta name="FILENAME" content="02.performance.fm">
     
<title>I/O Tuning</title>

</HEAD>

<BODY bgcolor="#FFFFFF" link="#000066" vlink="#0000ff" marginheight="0" topmargin="0">
<basefont=1>

<a name="120371"></a>

<div align="left">
<table border="0" cellspacing="0" cellpadding="0" width="100%">
 <tr>
   <td height="2" border="0" cellspacing="0" cellpadding="0" colspan="2" bgcolor="#000000">
   </td>
 </tr>
 <tr>
   <td height="15" width="50%" border="0" cellspacing="0" cellpadding="0" bgcolor="#cccccc">
    <img src="images/vxsymbol.gif" height="20" width="31" border="0" cellspacing="0" cellpadding="0" align="left">
     
     
   </td>
   <td align="right" width="50%" bgcolor="#cccccc">

<!-- Header Breadcrumbs -->
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="-2">&lt;&nbsp;<a href="02_performance_vxfs33_120295.html" title="Previous - Monitoring Free Space">Previous</a>&nbsp;&nbsp;|&nbsp;&nbsp;<a href="03_extent_att_vxfs35_122044.html" title="Next - Extent Attributes">Next</a>
&nbsp;&gt;&nbsp;&nbsp;</font>
   </td>
 </tr>

 <tr>
  <td bgcolor="#CCCCCC" colspan="2" height="15" align="right">
  <font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="1">Product:  
<a href="vxfs_splashpage_docCD_vxfs1_103036.vxhd" title="File System Guides title page ">File System Guides</a>
  &nbsp;&nbsp;
  </font>
  </td>
 </tr>
 <tr>
  <td bgcolor="#CCCCCC" colspan="2" height="15" align="right">
  <font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="1">Manual:  
<a href="vxfs_admin_titlepage_docCD_vxfs2_73700.vxhd" title="File System 4.1 Administrator's Guide title page ">File System 4.1 Administrator's Guide</a>
  &nbsp;&nbsp;
  </font>
  </td>
 </tr>


</table>
</div>

<div align="left">
 <table border="0" width="100%" bgcolor="#FFFFFF">
  <tr>
   <td width="100%" valign="top" bgcolor="#FFFFFF">

<!-- Start Main Topic Content -->
        <a name="startcontent"></a>
	
<h1><font face="Arial, Helvetica, Geneva, sansserif, sans serif, Verdana" size="+1" color="#000000"><a name="120371"></a> I/O Tuning</font>
</h1>
         
<hr noshade size="2" width="100%" color="#000000">
<img src="images/note.gif" alt="Note " height="16" width="16">&nbsp;
<font face="Arial, Helvetica, Geneva, sansserif, sans serif, Verdana" size="2" color="#000000">
<a name="120372"></a><b>Note  &nbsp;&nbsp;</b>
</font>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2" color="#000000">The tunables and the techniques described in this section work on a per file system basis. Use them judiciously based on the underlying device properties and characteristics of the applications that use the file system.
</font>
<hr noshade size="2" width="100%" color="#000000">
<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120373"></a>Performance of a file system can be enhanced by a suitable choice of I/O sizes and proper alignment of the I/O requests based on the requirements of the underlying special device. VxFS provides tools to tune the file systems.</font>
</p>




<h2><font face="Arial, Helvetica, Geneva, sansserif, sans serif, Verdana" size="3" color="#000000"><a name="120378"></a>Tuning VxFS I/O Parameters</font>
</h2>
 
<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120379"></a>VxFS provides a set of tunable I/O parameters that control some of its behavior. These I/O parameters are useful to help the file system adjust to striped or RAID-5 volumes that could yield performance superior to a single disk. Typically, data streaming applications that access large files see the largest benefit from tuning the file system. </font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="181796"></a>The file system queries VxVM to determine the geometry of the underlying volume and automatically sets the I/O parameters. The <font face="Courier New, Courier, sansserif, sans serif" size="2">mount</font> command also queries VxVM when the file system is mounted and downloads the I/O parameters.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120384"></a>If the default parameters are not acceptable or the file system is being used without VxVM, then the <font face="Courier New, Courier, sansserif, sans serif" size="2">/etc/vx/tunefstab</font> file can be used to set values for I/O parameters. The <font face="Courier New, Courier, sansserif, sans serif" size="2">mount</font> command reads the <font face="Courier New, Courier, sansserif, sans serif" size="2">/etc/vx/tunefstab</font> file and downloads any parameters specified for a file system. The <font face="Courier New, Courier, sansserif, sans serif" size="2">tunefstab</font> file overrides any values obtained from VxVM. While the file system is mounted, any I/O parameters can be changed using the <font face="Courier New, Courier, sansserif, sans serif" size="2">vxtunefs</font> command which can have tunables specified on the command line or can read them from the <font face="Courier New, Courier, sansserif, sans serif" size="2">/etc/vx/tunefstab</font> file. For more details, see the <font face="Courier New, Courier, sansserif, sans serif" size="2">vxtunefs</font>(1M) and <font face="Courier New, Courier, sansserif, sans serif" size="2">tunefstab</font>(4) manual pages. The <font face="Courier New, Courier, sansserif, sans serif" size="2">vxtunefs</font> command can be used to print the current values of the I/O parameters:</font>
</p>

<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120385"></a>&nbsp;#&nbsp;<font face="Courier New, Courier, sansserif, sans serif" size="2"><b>vxtunefs&nbsp;-p</b></font>&nbsp;<font face="Courier New, Courier, sansserif, sans serif" size="2"><b>mount_point</b></font>
</font>
<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120386"></a>The following is an example <font face="Courier New, Courier, sansserif, sans serif" size="2">tunefstab</font> file:</font>
</p>

<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120387"></a>&nbsp;/dev/vx/dsk/userdg/netbackup
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120388"></a>&nbsp;read_pref_io=128k,write_pref_io=128k,read_nstream=4,write_nstream=4
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120391"></a>&nbsp;/dev/vx/dsk/userdg/metasave
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120392"></a>&nbsp;read_pref_io=128k,write_pref_io=128k,read_nstream=4,write_nstream=4
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120393"></a>&nbsp;/dev/vx/dsk/userdg/solbuild
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120394"></a>&nbsp;read_pref_io=64k,write_pref_io=64k,read_nstream=4,write_nstream=4
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120395"></a>&nbsp;/dev/vx/dsk/userdg/solrelease
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120396"></a>&nbsp;read_pref_io=64k,write_pref_io=64k,read_nstream=4,write_nstream=4
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120397"></a>&nbsp;/dev/vx/dsk/userdg/solpatch
</font>
<br>
<font face="Courier New, Courier, sansserif, sans serif" size="2"><a name="120398"></a>&nbsp;read_pref_io=128k,write_pref_io=128k,read_nstream=4,write_nstream=4
</font>



<h2><font face="Arial, Helvetica, Geneva, sansserif, sans serif, Verdana" size="3" color="#000000"><a name="120405"></a>Tunable VxFS I/O Parameters</font>
</h2>
 
<a name="no_figure_caption0"> </a>
<Table width="100%" border="0" cellpadding="1" cellspacing="0" >
<tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219514"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">read_pref_io</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219516"></a>The preferred read request size. The file system uses this in conjunction with the <font face="Courier New, Courier, sansserif, sans serif" size="2">read_nstream</font> value to determine how much data to read ahead. The default value is 64K.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219519"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">write_pref_io</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219521"></a>The preferred write request size. The file system uses this in conjunction with the <font face="Courier New, Courier, sansserif, sans serif" size="2">write_nstream</font> value to determine how to do flush behind on writes. The default value is 64K.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219524"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">read_nstream</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219526"></a>The number of parallel read requests of size <font face="Courier New, Courier, sansserif, sans serif" size="2">read_pref_io</font> to have outstanding at one time. The file system uses the product of <font face="Courier New, Courier, sansserif, sans serif" size="2">read_nstream</font> multiplied by <font face="Courier New, Courier, sansserif, sans serif" size="2">read_pref_io</font> to determine its read ahead size. The default value for <font face="Courier New, Courier, sansserif, sans serif" size="2">read_nstream</font> is 1.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219529"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">write_nstream</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219531"></a>The number of parallel write requests of size <font face="Courier New, Courier, sansserif, sans serif" size="2">write_pref_io</font> to have outstanding at one time. The file system uses the product of <font face="Courier New, Courier, sansserif, sans serif" size="2">write_nstream</font> multiplied by <font face="Courier New, Courier, sansserif, sans serif" size="2">write_pref_io</font> to determine when to do flush behind on writes. The default value for <font face="Courier New, Courier, sansserif, sans serif" size="2">write_nstream</font> is 1.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219538"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">discovered_direct_iosz</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219541"></a>Any file I/O requests larger than the <font face="Courier New, Courier, sansserif, sans serif" size="2">discovered_direct_iosz</font> are handled as discovered direct I/O. A discovered direct I/O is unbuffered similar to direct I/O, but it does not require a synchronous commit of the inode when the file is extended or blocks are allocated. For larger I/O requests, the CPU time for copying the data into the page cache and the cost of using memory to buffer the I/O data becomes more expensive than the cost of doing the disk I/O. For these I/O requests, using discovered direct I/O is more efficient than regular I/O. The default value of this parameter is 256K.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219543"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">default_indir_<br>siz</font>e</font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219546"></a>On VxFS, files can have up to ten direct extents of variable size stored in the inode. Once these extents are used up, the file must use indirect extents which are a fixed size that is set when the file first uses indirect extents. These indirect extents are 8K by default. The file system does not use larger indirect extents because it must fail a write and return <font face="Courier New, Courier, sansserif, sans serif" size="2">ENOSPC</font> if there are no extents available that are the indirect extent size. For file systems with many large files, the 8K indirect extent size is too small. The files that get into indirect extents use many smaller extents instead of a few larger ones. By using this parameter, the default indirect extent size can be increased so large that files in indirects use fewer larger extents. The tunable <font face="Courier New, Courier, sansserif, sans serif" size="2">default_indir_size</font> should be used carefully. If it is set too large, then writes will fail when they are unable to allocate extents of the indirect extent size to a file. In general, the fewer and the larger the files on a file system, the larger the <font face="Courier New, Courier, sansserif, sans serif" size="2">default_indir_size</font> can be set. This parameter should generally be set to some multiple of the <font face="Courier New, Courier, sansserif, sans serif" size="2">read_pref_io</font> parameter. <font face="Courier New, Courier, sansserif, sans serif" size="2">default_indir_size</font> is not applicable on Version 4 disk layouts.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219549"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_keeptime</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219551"></a>Specifies the minimum amount of time, in seconds, that the VxFS file change log (FCL) keeps records in the log. When the oldest 8K block of FCL records have been kept longer than the value of <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_keeptime</font>, they are purged from the FCL and the extents nearest to the beginning of the FCL file are freed. This process is referred to as &quot;punching a hole.&quot; Holes are punched in the FCL file in 8K chunks. </font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219553"></a>If the  <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font> parameter is set, records are purged from the FCL if the amount of space allocated to the FCL exceeds <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font>, even if the elapsed time the records have been in the log is less than the value of <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_keeptime</font>. If the file system runs out of space before <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_keeptime</font> is reached, the FCL is deactivated.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219554"></a>Either or both of the  <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_keeptime</font> or  <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font> parameters must be set before the file change log can be activated. <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_keeptime</font> does not apply to disk layout Versions 1 through 5.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219557"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219559"></a>Specifies the maximum amount of space that can be allocated to the VxFS file change log (FCL). The FCL file is a sparse file that grows as changes occur in the file system. When the space allocated to the FCL file reaches the <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font> value, the oldest FCL records are purged from the FCL and the extents nearest to the beginning of the FCL file are freed. This process is referred to as &quot;punching a hole.&quot; Holes are punched in the FCL file in 8K chunks. If the file system runs out of space before  <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font> is reached, the FCL is deactivated.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219560"></a>Either or both of the  <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font> or  <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_keeptime</font> parameters must be set before the file change log can be activated. <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_maxalloc</font> does not apply to disk lay out Versions 1 through 5.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219563"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_winterval</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219565"></a>Specifies the time, in seconds, that must elapse before the VxFS   file change log (FCL) records a data overwrite, data extending write, or data truncate for a file. The ability to limit the number of repetitive FCL records for continuous writes to the same file is important for file system performance and for applications processing the FCL. <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_winterval</font> is best set to an interval less than the shortest interval between reads of the FCL by any application. This way all applications using the FCL can be assured of finding at least one FCL record for any file experiencing continuous data changes.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219566"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_winterval</font> is enforced for all  files  in  the  file system. Each  file maintains its own time stamps, and the elapsed time between FCL records is per file.  This elapsed  time can be overridden using the VxFS FCL sync public API (see the <font face="Courier New, Courier, sansserif, sans serif" size="2">vxfs_fcl_sync</font>(3) manual page). <font face="Courier New, Courier, sansserif, sans serif" size="2">fcl_winterval</font> does not apply to disk layout Versions  1 through 5.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219569"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">hsm_write_<br>prealloc</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219572"></a>For a file managed by a hierarchical storage management (HSM) application, <font face="Courier New, Courier, sansserif, sans serif" size="2">hsm_write_prealloc</font> preallocates disk blocks before data is migrated back into the file system. An HSM application usually migrates the data back through a series of writes to the file, each of which allocates a few blocks. By setting <font face="Courier New, Courier, sansserif, sans serif" size="2">hsm_write_prealloc</font> (<font face="Courier New, Courier, sansserif, sans serif" size="2">hsm_write_prealloc</font>=1), a sufficient number of disk blocks are allocated on the first write to the empty file so that no disk block allocation is required for subsequent writes. This improves the write performance during migration.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219573"></a>The <font face="Courier New, Courier, sansserif, sans serif" size="2">hsm_write_prealloc</font> parameter is implemented outside of the DMAPI specification, and its usage has limitations depending on how the space within an HSM-controlled file is managed. It is advisable to use <font face="Courier New, Courier, sansserif, sans serif" size="2">hsm_write_prealloc</font> only when recommended by the HSM application controlling the file system.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219575"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">initial_extent_<br>size</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219578"></a>Changes the default initial extent size. VxFS determines, based on the first write to a new file, the size of the first extent to be allocated to the file. Normally the first extent is the smallest power of 2 that is larger than the size of the first write. If that power of 2 is less than 8K, the first extent allocated is 8K. After the initial extent, the file system increases the size of subsequent extents (see <font face="Courier New, Courier, sansserif, sans serif" size="2">max_seqio_extent_size</font>) with each allocation. Since most applications write to files using a buffer size of 8K or less, the increasing extents start doubling from a small initial extent. <font face="Courier New, Courier, sansserif, sans serif" size="2">initial_extent_size</font> can change the default initial extent size to be larger, so the doubling policy will start from a much larger initial size and the file system will not allocate a set of small extents at the start of file. Use this parameter only on file systems that will have a very large average file size. On these file systems it will result in fewer extents per file and less fragmentation. <font face="Courier New, Courier, sansserif, sans serif" size="2">initial_extent_size</font> is measured in file system blocks.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219581"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">max_buf_data_size</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219583"></a>The maximum buffer size allocated for file data; either 8K bytes or 64K bytes. Use the larger value for workloads where large reads/writes are performed sequentially. Use the smaller value on workloads where the I/O is random or is done in small chunks. 8K bytes is the default value.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219586"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">inode_aging_count</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219588"></a>Specifies the maximum number of inodes to place on an inode aging list. Inode aging is used in conjunction with file system Storage Checkpoints to allow quick restoration of large,   recently deleted files. The aging list is maintained in first-in-first-out (fifo) order up to maximum number of inodes specified by <font face="Courier New, Courier, sansserif, sans serif" size="2">inode_aging_count</font>. As newer inodes are placed on the list, older inodes are removed to complete their aging process. For best performance, it is advisable to age only a limited number of larger files before completion of the removal process. The default maximum number of inodes to age is 2048.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219591"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">inode_aging_size</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219593"></a> Specifies the minimum size to qualify a deleted inode for inode aging. Inode aging is used in conjunction with file system Storage Checkpoints to allow quick restoration of large, recently deleted files. For best performance, it is advisable to age only a limited number of larger files before completion of the removal process. Setting the size too low can push larger file inodes out of the aging queue to make room for newly removed smaller file inodes.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219596"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">max_direct_iosz</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219598"></a>The maximum size of a direct I/O request that will be issued by the file system. If a larger I/O request comes in, then it is broken up into <font face="Courier New, Courier, sansserif, sans serif" size="2">max_direct_iosz</font> chunks. This parameter defines how much memory an I/O request can lock at once, so it should not be set to more than 20 percent of memory.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219601"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">max_diskq</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219603"></a>Limits the maximum disk queue generated by a single file. When the file system is flushing data for a file and the number of buffers being flushed exceeds <font face="Courier New, Courier, sansserif, sans serif" size="2">max_diskq</font>, processes will block until the amount of data being flushed decreases. Although this doesn't limit the actual disk queue, it prevents flushing processes from making the system unresponsive. The default value is 1 MB.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219607"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">max_seqio_extent_size</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219610"></a>Increases or decreases the maximum size of an extent. When the file system is following its default allocation policy for sequential writes to a file, it allocates an initial extent which is large enough for the first write to the file. When additional extents are allocated, they are progressively larger (the algorithm tries to double the size of the file with each new extent) so each extent can hold several writes worth of data. This is done to reduce the total number of extents in anticipation of continued sequential writes. When the file stops being written, any unused space is freed for other files to use. Normally this allocation stops increasing the size of extents at 2048 blocks which prevents one file from holding too much unused space. <font face="Courier New, Courier, sansserif, sans serif" size="2">max_seqio_extent_size</font> is measured in file system blocks.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219623"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">qio_cache_enable</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219625"></a>Enables or disables caching on Quick I/O files. The default behavior is to disable caching. To enable caching, set <font face="Courier New, Courier, sansserif, sans serif" size="2">qio_cache_enable</font> to 1. On systems with large memories, the database cannot always use all of the memory as a cache. By enabling file system caching as a second level cache, performance may be improved. If the database is performing sequential scans of tables, the scans may run faster by enabling file system caching so the file system will perform aggressive read-ahead on the files.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219683"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">read_ahead</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219686"></a>The default for all VxFS read operations is to perform sequential read ahead. You can specify the <font face="Courier New, Courier, sansserif, sans serif" size="2">read_ahead</font> cache advisory to implement the VxFS <em>enhanced read ahead functionality</em>. This allows read aheads to detect more elaborate patterns (such as increasing or decreasing read offsets or multithreaded file accesses) in addition to simple sequential reads. You can specify the following values for <font face="Courier New, Courier, sansserif, sans serif" size="2">read_ahead</font>:</font>
</p>

<blockquote>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219687"></a>0---Disables read ahead functionality</font>
</blockquote>
<blockquote>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219688"></a>1---Retains traditional sequential read ahead behavior</font>
</blockquote>
<blockquote>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219689"></a>2---Enables enhanced read ahead for all reads</font>
</blockquote>
<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219690"></a>The default is 1---VxFS detects only sequential patterns.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219691"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">read_ahead</font> detects patterns on a per-thread basis, up to a maximum determined by <font face="Courier New, Courier, sansserif, sans serif" size="2">vx_era_nthreads</font> parameter. The default number of threads is 5, but you can change the default value by setting the <font face="Courier New, Courier, sansserif, sans serif" size="2">vx_era_nthreads</font> parameter in the <font face="Courier New, Courier, sansserif, sans serif" size="2">/etc/system</font> configuration file.</font>
</p>

</td></tr><tr valign="top"><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219694"></a><font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font></font>
</p>

</td><td><p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219696"></a>The <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> parameter is useful in special situations where a computer system has a combination of a large amount of memory and slow storage devices. In this configuration, sync operations (such as <font face="Courier New, Courier, sansserif, sans serif" size="2">fsync()</font>) may take long enough to complete that a system appears to hang. This behavior occurs because the file system is creating <em>dirty buffers</em> (in-memory updates) faster than they can be asynchronously flushed to disk without slowing system performance.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219697"></a>Lowering the value of <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> limits the number of dirty buffers per file that a file system will generate before flushing the buffers to disk. After the number of dirty buffers for a file reaches the <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> threshold, the file system starts flushing buffers to disk even if free memory is still available.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219698"></a>The default value of <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> is zero, which puts no limit on the number of dirty buffers per file. If non-zero, VxFS limits the number of dirty buffers per file to <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> buffers.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219699"></a>The default value typically generates a large number of dirty buffers, but maintains fast user writes. Depending on the speed of the storage device, if you lower <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font>, user write performance may suffer, but the number of dirty buffers is limited, so sync operations will complete much faster.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="219700"></a>Because lowering <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> may in some cases delay write requests (for example, lowering <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> may increase the file disk queue to the <font face="Courier New, Courier, sansserif, sans serif" size="2">max_diskq</font> value, delaying user writes until the disk queue decreases), it is advisable not to change the value of <font face="Courier New, Courier, sansserif, sans serif" size="2">write_throttle</font> unless your system has a combination of large physical memory and slow storage devices.</font>
</p>

</td></tr>
</Table>


<Table width="100%" border="0" cellpadding="1" cellspacing="0">
<tr><td>
</Table>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120559"></a>If the file system is being used with VxVM, it is advisable to let the VxFS I/O parameters get set to default values based on the volume geometry.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120560"></a>If the file system is being used with a hardware disk array or volume manager other than VxVM, try to align the parameters to match the geometry of the logical disk. With striping or RAID-5, it is common to set <font face="Courier New, Courier, sansserif, sans serif" size="2">read_pref_io</font> to the stripe unit size and <font face="Courier New, Courier, sansserif, sans serif" size="2">read_nstream</font> to the number of columns in the stripe. For striped arrays, use the same values for <font face="Courier New, Courier, sansserif, sans serif" size="2">write_pref_io</font> and <font face="Courier New, Courier, sansserif, sans serif" size="2">write_nstream</font>, but for RAID-5 arrays, set <font face="Courier New, Courier, sansserif, sans serif" size="2">write_pref_io</font> to the full stripe size and <font face="Courier New, Courier, sansserif, sans serif" size="2">write_nstream</font> to 1.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120561"></a>For an application to do efficient disk I/O, it should issue read requests that are equal to the product of <font face="Courier New, Courier, sansserif, sans serif" size="2">read_nstream</font> multiplied by <font face="Courier New, Courier, sansserif, sans serif" size="2">read_pref_io</font>. Generally, any multiple or factor of <font face="Courier New, Courier, sansserif, sans serif" size="2">read_nstream</font> multiplied by <font face="Courier New, Courier, sansserif, sans serif" size="2">read_pref_io</font> should be a good size for performance. For writing, the same rule of thumb applies to the <font face="Courier New, Courier, sansserif, sans serif" size="2">write_pref_io</font> and <font face="Courier New, Courier, sansserif, sans serif" size="2">write_nstream</font> parameters. When tuning a file system, the best thing to do is try out the tuning parameters under a real life workload.</font>
</p>

<p>
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="2"><a name="120562"></a>If an application is doing sequential I/O to large files, it should try to issue requests larger than the <font face="Courier New, Courier, sansserif, sans serif" size="2">discovered_direct_iosz</font>. This causes the I/O requests to be performed as discovered direct I/O requests, which are unbuffered like direct I/O but do not require synchronous inode updates when extending the file. If the file is larger than can fit in the cache, using unbuffered I/O avoids removing useful data out of the cache and lessens CPU overhead.</font>
</p>


<!-- End Main Topic Content -->

   </td>
  </tr>
 </table>
</div>

<div>
<table border="0" cellspacing="0" cellpadding="0" width="100%">
 <tr>
  <td height="18" border="0" cellspacing="0" cellpadding="0" colspan="2" bgcolor="#FFFFFF">
  </td>
 </tr>
  <td align="left"><font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="-2">&nbsp;&#94;&nbsp;<a href="02_performance_vxfs34_120371.html" title="I/O Tuning">Return to Top</a></font>

  </td>
  <td align="right"><font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="-2">&lt;&nbsp;<a href="02_performance_vxfs33_120295.html" title="Previous - Monitoring Free Space">Previous</a>&nbsp;&nbsp;|&nbsp;&nbsp;<a href="03_extent_att_vxfs35_122044.html" title="Next - Extent Attributes">Next</a>
&nbsp;&gt;&nbsp;&nbsp;</font>
  </td>
 <tr>
   <td height="9" border="0" cellspacing="0" cellpadding="0" colspan="2" bgcolor="#FFFFFF">
   </td>
 </tr>

<!-- Footer Breadcrumbs -->
<tr>
<td bgcolor="#FFFFFF" colspan="2" height="15" align="right">
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="1">Product:  
<a href="vxfs_splashpage_docCD_vxfs1_103036.vxhd" title="File System Guides title page ">File System Guides</a>&nbsp;&nbsp;</font></td>
</tr>

<tr>
<td bgcolor="#FFFFFF" colspan="2" height="15" align="right">
<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="1">Manual:  
<a href="vxfs_admin_titlepage_docCD_vxfs2_73700.vxhd" title="File System 4.1 Administrator's Guide title page ">File System 4.1 Administrator's Guide</a>&nbsp;&nbsp;</font></td>
</tr>


 <tr>
  <td colspan="2" align="left"><font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" color="#000000" align="right" size="-2">
<em>VERITAS Software Corporation
</em></font><br>

<font face="Verdana, Arial, Helvetica, Geneva, sansserif, sans serif" size="-2">www.veritas.com</font><br>
  </td>
 </tr>
</table>
</div>

</BODY>
  <!--
    *************************************************************************************
    * Copyright (c) 2005 VERITAS Software Corporation. All rights reserved. *

    VERITAS, VERITAS SOFTWARE, the VERITAS Logo, and VERITAS Software Documentation 
    are trademarks or registered trademarks of VERITAS Software Corporation 
    or its affiliates in the U.S. and other countries. Other names may be 
    trademarks of their respective owners. 

    Apache Jakarta Lucene:
    Apache Software License
    Version 1.1

    Copyright (c) 2000 The Apache Software Foundation. All rights reserved.

    Redistribution and use in source and binary forms, with or without modification, 
    are permitted provided that the following conditions are met:

    1. Redistributions of source code must retain the above copyright notice, this 
    list of conditions and the following disclaimer.

    2. Redistributions in binary form must reproduce the above copyright notice, this 
    list of conditions and the following disclaimer in the documentation and/or other 
    materials provided with the distribution.

    3. The end-user documentation included with the redistribution, if any, must include 
    the following acknowledgment: "This product includes software developed by the Apache 
    Software Foundation (http://www.apache.org/)." Alternately, this acknowledgment may 
    appear in the software itself, if and wherever such third-party acknowledgments 
    normally appear.

    4. The names "Apache" and "Apache Software Foundation" must not be used to endorse or 
    promote products derived from this software without prior written permission. For 
    written permission, please contact apache@apache.org.

    5. Products derived from this software may not be called "Apache", nor may "Apache" 
    appear in their name, without prior written permission of the Apache Software Foundation.

    THIS SOFTWARE IS PROVIDED "AS IS" AND ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, 
    BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A 
    PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION 
    OR ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, 
    OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 
    OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND 
    ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING 
    NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF 
    ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

    This software consists of voluntary contributions made by many individuals on behalf 
    of the Apache Software Foundation. For more information on the Apache Software Foundation, 
    please see <http://www.apache.org/>. Portions of this software are based upon public domain 
    software originally written at the National Center for Supercomputing Applications, University 
    of Illinois, Urbana-Champaign.
    
    This product includes WebWindow Software developed by Horst Heistermann.
    *************************************************************************************
  -->
</HTML>
