diff options
Diffstat (limited to 'docs/htmldocs/speed.html')
-rw-r--r-- | docs/htmldocs/speed.html | 140 |
1 files changed, 140 insertions, 0 deletions
diff --git a/docs/htmldocs/speed.html b/docs/htmldocs/speed.html new file mode 100644 index 0000000000..47f19abb70 --- /dev/null +++ b/docs/htmldocs/speed.html @@ -0,0 +1,140 @@ +<html><head><meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"><title>Chapter 39. Samba Performance Tuning</title><link rel="stylesheet" href="samba.css" type="text/css"><meta name="generator" content="DocBook XSL Stylesheets V1.60.1"><link rel="home" href="samba-doc.html" title="SAMBA Project Documentation"><link rel="up" href="Appendixes.html" title="Part VI. Appendixes"><link rel="previous" href="Other-Clients.html" title="Chapter 38. Samba and other CIFS clients"><link rel="next" href="DNSDHCP.html" title="Chapter 40. DNS and DHCP Configuration Guide"></head><body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="3" align="center">Chapter 39. Samba Performance Tuning</th></tr><tr><td width="20%" align="left"><a accesskey="p" href="Other-Clients.html">Prev</a> </td><th width="60%" align="center">Part VI. Appendixes</th><td width="20%" align="right"> <a accesskey="n" href="DNSDHCP.html">Next</a></td></tr></table><hr></div><div class="chapter" lang="en"><div class="titlepage"><div><div><h2 class="title"><a name="speed"></a>Chapter 39. Samba Performance Tuning</h2></div><div><div class="author"><h3 class="author"><span class="firstname">Paul</span> <span class="surname">Cochrane</span></h3><div class="affiliation"><span class="orgname">Dundee Limb Fitting Centre<br></span><div class="address"><p><tt class="email"><<a href="mailto:paulc@dth.scot.nhs.uk">paulc@dth.scot.nhs.uk</a>></tt></p></div></div></div></div><div><div class="author"><h3 class="author"><span class="firstname">Jelmer</span> <span class="othername">R.</span> <span class="surname">Vernooij</span></h3><div class="affiliation"><span class="orgname">The Samba Team<br></span><div class="address"><p><tt class="email"><<a href="mailto:jelmer@samba.org">jelmer@samba.org</a>></tt></p></div></div></div></div><div><div class="author"><h3 class="author"><span class="firstname">John</span> <span class="othername">H.</span> <span class="surname">Terpstra</span></h3><div class="affiliation"><span class="orgname">Samba Team<br></span><div class="address"><p><tt class="email"><<a href="mailto:jht@samba.org">jht@samba.org</a>></tt></p></div></div></div></div></div><div></div></div><div class="toc"><p><b>Table of Contents</b></p><dl><dt><a href="speed.html#id2964209">Comparisons</a></dt><dt><a href="speed.html#id2964253">Socket options</a></dt><dt><a href="speed.html#id2964343">Read size</a></dt><dt><a href="speed.html#id2964393">Max xmit</a></dt><dt><a href="speed.html#id2964453">Log level</a></dt><dt><a href="speed.html#id2964483">Read raw</a></dt><dt><a href="speed.html#id2964567">Write raw</a></dt><dt><a href="speed.html#id2964630">Slow Logins</a></dt><dt><a href="speed.html#id2964659">Client tuning</a></dt><dt><a href="speed.html#id2964684">Samba performance problem due changing kernel</a></dt><dt><a href="speed.html#id2964716">Corrupt tdb Files</a></dt></dl></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964209"></a>Comparisons</h2></div></div><div></div></div><p> +The Samba server uses TCP to talk to the client. Thus if you are +trying to see if it performs well you should really compare it to +programs that use the same protocol. The most readily available +programs for file transfer that use TCP are ftp or another TCP based +SMB server. +</p><p> +If you want to test against something like a NT or WfWg server then +you will have to disable all but TCP on either the client or +server. Otherwise you may well be using a totally different protocol +(such as NetBEUI) and comparisons may not be valid. +</p><p> +Generally you should find that Samba performs similarly to ftp at raw +transfer speed. It should perform quite a bit faster than NFS, +although this very much depends on your system. +</p><p> +Several people have done comparisons between Samba and Novell, NFS or +WinNT. In some cases Samba performed the best, in others the worst. I +suspect the biggest factor is not Samba vs some other system but the +hardware and drivers used on the various systems. Given similar +hardware Samba should certainly be competitive in speed with other +systems. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964253"></a>Socket options</h2></div></div><div></div></div><p> +There are a number of socket options that can greatly affect the +performance of a TCP based server like Samba. +</p><p> +The socket options that Samba uses are settable both on the command +line with the <tt class="option">-O</tt> option, or in the <tt class="filename">smb.conf</tt> file. +</p><p> +The <a class="indexterm" name="id2964287"></a><i class="parameter"><tt>socket options</tt></i> section of the <tt class="filename">smb.conf</tt> manual page describes how +to set these and gives recommendations. +</p><p> +Getting the socket options right can make a big difference to your +performance, but getting them wrong can degrade it by just as +much. The correct settings are very dependent on your local network. +</p><p> +The socket option TCP_NODELAY is the one that seems to make the +biggest single difference for most networks. Many people report that +adding <a class="indexterm" name="id2964323"></a><i class="parameter"><tt>socket options</tt></i> = TCP_NODELAY doubles the read +performance of a Samba drive. The best explanation I have seen for this is +that the Microsoft TCP/IP stack is slow in sending tcp ACKs. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964343"></a>Read size</h2></div></div><div></div></div><p> +The option <a class="indexterm" name="id2964353"></a><i class="parameter"><tt>read size</tt></i> affects the overlap of disk +reads/writes with network reads/writes. If the amount of data being +transferred in several of the SMB commands (currently SMBwrite, SMBwriteX and +SMBreadbraw) is larger than this value then the server begins writing +the data before it has received the whole packet from the network, or +in the case of SMBreadbraw, it begins writing to the network before +all the data has been read from disk. +</p><p> +This overlapping works best when the speeds of disk and network access +are similar, having very little effect when the speed of one is much +greater than the other. +</p><p> +The default value is 16384, but very little experimentation has been +done yet to determine the optimal value, and it is likely that the best +value will vary greatly between systems anyway. A value over 65536 is +pointless and will cause you to allocate memory unnecessarily. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964393"></a>Max xmit</h2></div></div><div></div></div><p> + At startup the client and server negotiate a <i class="parameter"><tt>maximum transmit</tt></i> size, +which limits the size of nearly all SMB commands. You can set the +maximum size that Samba will negotiate using the <a class="indexterm" name="id2964414"></a><i class="parameter"><tt>max xmit</tt></i> option +in <tt class="filename">smb.conf</tt>. Note that this is the maximum size of SMB requests that +Samba will accept, but not the maximum size that the *client* will accept. +The client maximum receive size is sent to Samba by the client and Samba +honours this limit. +</p><p> +It defaults to 65536 bytes (the maximum), but it is possible that some +clients may perform better with a smaller transmit unit. Trying values +of less than 2048 is likely to cause severe problems. +</p><p> +In most cases the default is the best option. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964453"></a>Log level</h2></div></div><div></div></div><p> +If you set the log level (also known as <a class="indexterm" name="id2964463"></a><i class="parameter"><tt>debug level</tt></i>) higher than 2 +then you may suffer a large drop in performance. This is because the +server flushes the log file after each operation, which can be very +expensive. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964483"></a>Read raw</h2></div></div><div></div></div><p> +The <a class="indexterm" name="id2964492"></a><i class="parameter"><tt>read raw</tt></i> operation is designed to be an optimised, low-latency +file read operation. A server may choose to not support it, +however. and Samba makes support for <a class="indexterm" name="id2964510"></a><i class="parameter"><tt>read raw</tt></i> optional, with it +being enabled by default. +</p><p> +In some cases clients don't handle <a class="indexterm" name="id2964528"></a><i class="parameter"><tt>read raw</tt></i> very well and actually +get lower performance using it than they get using the conventional +read operations. +</p><p> +So you might like to try <a class="indexterm" name="id2964549"></a><i class="parameter"><tt>read raw</tt></i> = no and see what happens on your +network. It might lower, raise or not affect your performance. Only +testing can really tell. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964567"></a>Write raw</h2></div></div><div></div></div><p> +The <a class="indexterm" name="id2964577"></a><i class="parameter"><tt>write raw</tt></i> operation is designed to be an optimised, low-latency +file write operation. A server may choose to not support it, +however. and Samba makes support for <a class="indexterm" name="id2964594"></a><i class="parameter"><tt>write raw</tt></i> optional, with it +being enabled by default. +</p><p> +Some machines may find <a class="indexterm" name="id2964613"></a><i class="parameter"><tt>write raw</tt></i> slower than normal write, in which +case you may wish to change this option. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964630"></a>Slow Logins</h2></div></div><div></div></div><p> +Slow logins are almost always due to the password checking time. Using +the lowest practical <a class="indexterm" name="id2964641"></a><i class="parameter"><tt>password level</tt></i> will improve things. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964659"></a>Client tuning</h2></div></div><div></div></div><p> +Often a speed problem can be traced to the client. The client (for +example Windows for Workgroups) can often be tuned for better TCP +performance. Check the sections on the various clients in +<a href="Other-Clients.html" title="Chapter 38. Samba and other CIFS clients">Samba and Other Clients</a>. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964684"></a>Samba performance problem due changing kernel</h2></div></div><div></div></div><p> +Hi everyone. I am running Gentoo on my server and samba 2.2.8a. Recently +I changed kernel version from linux-2.4.19-gentoo-r10 to +linux-2.4.20-wolk4.0s. And now I have performance issue with samba. Ok +many of you will probably say that move to vanilla sources...well I tried +it too and it didn't work. I have 100mb LAN and two computers (linux + +Windows2000). Linux server shares directory with DivX files, client +(windows2000) plays them via LAN. Before when I was running 2.4.19 kernel +everything was fine, but now movies freezes and stops...I tried moving +files between server and Windows and it's terribly slow. +</p><p> +Grab mii-tool and check the duplex settings on the NIC. +My guess is that it is a link layer issue, not an application +layer problem. Also run ifconfig and verify that the framing +error, collisions, etc... look normal for ethernet. +</p></div><div class="sect1" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="id2964716"></a>Corrupt tdb Files</h2></div></div><div></div></div><p> +Well today it happened, Our first major problem using samba. +Our samba PDC server has been hosting 3 TB of data to our 500+ users +[Windows NT/XP] for the last 3 years using samba, no problem. +But today all shares went SLOW; very slow. Also the main smbd kept +spawning new processes so we had 1600+ running smbd's (normally we avg. 250). +It crashed the SUN E3500 cluster twice. After a lot of searching I +decided to <b class="command">rm /var/locks/*.tdb</b>. Happy again. +</p><p> +Q1) Is there any method of keeping the *.tdb files in top condition or +how to early detect corruption? +</p><p> +A1) Yes, run <b class="command">tdbbackup</b> each time after stopping nmbd and before starting nmbd. +</p><p> +Q2) What I also would like to mention is that the service latency seems +a lot lower then before the locks cleanup, any ideas on keeping it top notch? +</p><p> +A2) Yes! Same answer as for Q1! +</p></div></div><div class="navfooter"><hr><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="Other-Clients.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="Appendixes.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="DNSDHCP.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">Chapter 38. Samba and other CIFS clients </td><td width="20%" align="center"><a accesskey="h" href="samba-doc.html">Home</a></td><td width="40%" align="right" valign="top"> Chapter 40. DNS and DHCP Configuration Guide</td></tr></table></div></body></html> |