<chapter id="server">
<title>Server — How the server works</title>
<simplesect>
<para>The term “server” is ambiguous, because it has at least
two different meanings: it can refer to a powerful computer which offers
services to users on a network, or it can refer to a CPU process designed
to receive network requests.</para>
<para>In Subversion, however, the <firstterm>server</firstterm> is just a
set of libraries that implements <firstterm>repositories</firstterm> and
makes them available to other programs. No networking is
required.</para>
<para>There are two main libraries: the <firstterm>Subversion
Filesystem</firstterm> library, and the <firstterm>Subversion
Repository</firstterm> library.</para>
</simplesect>
<sect1 id="server.fs">
<title>Filesystem</title>
<sect2 id="server.fs.overview">
<title>Filesystem Overview</title>
<itemizedlist mark="bullet">
<listitem><para><emphasis role="bold">Requires:</emphasis>
<itemizedlist mark="minus">
<listitem><para>some writable disk space</para></listitem>
<listitem><para>(for now) Berkeley DB library</para></listitem>
</itemizedlist>
</para></listitem>
<listitem><para><emphasis role="bold">Provides:</emphasis>
<itemizedlist mark="minus">
<listitem><para>a repository for storing files</para></listitem>
<listitem><para>concurrent client transactions</para></listitem>
<listitem><para>enforcement of user & group permissions
[someday, not yet]</para></listitem>
</itemizedlist>
</para></listitem>
</itemizedlist>
<para>This library implements a hierarchical filesystem which supports
atomic changes to directory trees, and records a complete history of
the changes. In addition to recording changes to file and directory
contents, the Subversion Filesystem records changes to file meta-data
(see discussion of <firstterm>properties</firstterm> in <xref
linkend="model"/>).</para>
</sect2>
<sect2 id="server.fs.api">
<title>API</title>
<para> There are two main files that describe the Subversion
filesystem.</para>
<para>First, read the section below (<xref linkend="server.fs.struct"/>)
for a general overview of how the filesystem works.</para>
<para>Once you've done this, read Jim Blandy's own structural overview,
which explains how nodes and revisions are organized (among other
things) in the filesystem implementation:
<filename>subversion/libsvn_fs/structure</filename>.</para>
<para>Finally, read the well-documented API in
<filename>subversion/include/svn_fs.h</filename>.</para>
</sect2>
<sect2 id="server.fs.struct">
<title>Repository Structure</title>
<sect3 id="server.fs.struct.schema">
<title>Schema</title>
<para>
To begin, please be sure that you're already casually familiar with
Subversion's ideas of files, directories, and revision histories. If
not, see <xref linkend="model"/>. We can now offer precise,
technical descriptions of the terms introduced there.</para>
<!-- This is taken from jimb's very first Subversion spec! -->
<screen>
A <firstterm>text string</firstterm> is a string of Unicode characters which is
canonically decomposed and ordered, according to the rules described in the
Unicode standard.
A <firstterm>string of bytes</firstterm> is what you'd expect.
A <firstterm>property list</firstterm> is an unordered list of properties. A
<firstterm>property</firstterm> is a pair
<literal>(<replaceable>name</replaceable>,
<replaceable>value</replaceable>)</literal>, where
<replaceable>name</replaceable> is a text string, and
<replaceable>value</replaceable> is a string of bytes. No two properties in a
property list have the same name.
A <firstterm>file</firstterm> is a property list and a string of bytes.
A <firstterm>node</firstterm> is either a file or a directory. (We define a
directory below.) Nodes are distinguished unions — you can always tell
whether a node is a file or a directory.
A <firstterm>node table</firstterm> is an array mapping some set of positive
integers, called <firstterm>node numbers</firstterm>, onto
<firstterm>nodes</firstterm>. If a node table maps some number
<replaceable>i</replaceable> to some node <replaceable>n</replaceable>, then
<replaceable>i</replaceable> is a <firstterm>valid node number</firstterm> in
that table, and <firstterm>node</firstterm> <replaceable>i</replaceable>is
<replaceable>n</replaceable>. Otherwise, <replaceable>i</replaceable> is an
<firstterm>invalid node number</firstterm> in that table.
A <firstterm>directory entry</firstterm> is a triple
<literal>(<replaceable>name</replaceable>, <replaceable>props</replaceable>,
<replaceable>node</replaceable>)</literal>, where
<replaceable>name</replaceable> is a text string,
<replaceable>props</replaceable> is a property list, and
<replaceable>node</replaceable> is a node number.
A <firstterm>directory</firstterm> is an unordered list of directory entries,
and a property list.
A <firstterm>revision</firstterm> is a node number and a property list.
A <firstterm>history</firstterm> is an array of revisions, indexed by a
contiguous range of non-negative integers containing 0.
A <firstterm>repository</firstterm> consists of node table and a history.
</screen>
<!-- Some definitions: we say that a node @var{n} is a @dfn{direct
child} of a directory @var{d} iff @var{d} contains a directory entry
whose node number is @var{n}. A node @var{n} is a @dfn{child} of a
directory @var{d} iff @var{n} is a direct child of @var{d}, or if there
exists some directory @var{e} which is a direct child of @var{d}, and
@var{n} is a child of @var{e}. Given this definition of ``direct
child'' and ``child,'' the obvious definitions of ``direct parent'' and
``parent'' hold.
In these restrictions, let @var{r} be any repository. When we refer,
implicitly or explicitly, to a node table without further
clarification, we mean @var{r}'s node table. Thus, if we refer to ``a
valid node number'' without specifying the node table in which it is
valid, we mean ``a valid node number in @var{r}'s node table''.
Similarly for @var{r}'s history. -->
<para>Now that we've explained the form of the data, we make some
restrictions on that form.</para>
<para><emphasis role="bold">Every revision has a root
directory.</emphasis> Every revision's node number is a valid node
number, and the node it refers to is always a directory. We call
this the revision's <firstterm>root directory</firstterm>.</para>
<para><emphasis role="bold">Revision 0 always contains an empty root
directory.</emphasis> This baseline makes it easy to check out
whole projects from the repository.</para>
<para><emphasis role="bold">Directories contain only valid
links.</emphasis> Every directory entry's
<replaceable>node</replaceable> is a valid node number.</para>
<para><emphasis role="bold">Directory entries can be identified by
name.</emphasis> For any directory <replaceable>d</replaceable>,
every directory entry in <replaceable>d</replaceable> has a distinct
name.</para>
<para><emphasis role="bold">There are no cycles of
directories.</emphasis> No node is its own child.</para>
<para><emphasis role="bold">Directories can have more than one
parent.</emphasis> The Unix file system does not allow more than
one hard link to a directory, but Subversion does allow the analogous
situation. Thus, the directories in a Subversion repository form a
directed acyclic graph (<firstterm>DAG</firstterm>), not a tree.
However, it would be distracting and unhelpful to replace the
familiar term “directory tree” with the unfamiliar term
“directory DAG”, so we still call it a “directory
tree” here.</para>
<para><emphasis role="bold">There are no dead nodes.</emphasis> Every
node is a child of some revision's root directory.</para>
<!-- </jimb> -->
</sect3>
<sect3 id="server.fs.bubble-up">
<title>Bubble-Up Method</title>
<para>This section provides a conversational explanation of how the
repository actually stores and revisions file trees. It's not
critical knowledge for a programmer using the Subversion Filesystem
API, but most people probably still want to know what's going on
“under the hood” of the repository.</para>
<para>Suppose we have a new project, at revision 1, looking like this
(using CVS syntax):</para>
<programlisting>
prompt$ svn checkout myproj
U myproj/
U myproj/B
U myproj/A
U myproj/A/fish
U myproj/A/fish/tuna
prompt$
</programlisting>
<para>Only the file <filename>tuna</filename> is a regular file,
everything else in myproj is a directory.</para>
<para>Let's see what this looks like as an abstract data structure in
the repository, and how that structure works in various operations
(such as update, commit, and branch).</para>
<para>In the diagrams that follow, lines represent parent-to-child
connections in a directory hierarchy. Boxes are "nodes". A node is
either a file or a directory – a letter in the upper left
indicates which kind. A file node has a byte-string for its content,
whereas directory nodes have a list of dir_entries, each pointing to
another node.</para>
<para>Parent-child links go both ways (i.e., a child knows who all its
parents are), but a node's name is stored only in its parent, because
a node with multiple parents may have different names in different
parents.</para>
<para>At the top of the repository is an array of revision numbers,
stretching off to infinity. Since the project is at revision 1, only
index 1 points to anything; it points to the root node of revision 1
of the project:</para>
<programlisting>
( myproj's revision array )
______________________________________________________
|___1_______2________3________4________5_________6_____...
|
|
___|_____
|D |
| |
| A | /* Two dir_entries, `A' and `B'. */
| \ |
| B \ |
|__/___\__|
/ \
| \
| \
___|___ ___\____
|D | |D |
| | | |
| | | fish | /* One dir_entry, `fish'. */
|_______| |___\____|
\
\
___\____
|D |
| |
| tuna | /* One dir_entry, `tuna'. */
|___\____|
\
\
___\____
|F |
| |
| | /* (Contents of tuna not shown.) */
|________|
</programlisting>
<para>What happens when we modify <filename>tuna</filename> and commit?
First, we make a new <filename>tuna</filename> node, containing the
latest text. The new node is not connected to anything yet, it's
just hanging out there in space:</para>
<programlisting>
________
|F |
| |
| |
|________|
</programlisting>
<para>Next, we create a <emphasis>new</emphasis> revision of its parent
directory:</para>
<programlisting>
________
|D |
| |
| tuna |
|___\____|
\
\
___\____
|F |
| |
| |
|________|
</programlisting>
<para>We continue up the line, creating a new revision of the next
parent directory:</para>
<programlisting>
________
|D |
| |
| fish |
|___\____|
\
\
___\____
|D |
| |
| tuna |
|___\____|
\
\
___\____
|F |
| |
| |
|________|
</programlisting>
<para>Now it gets more tricky: we need to create a new revision of the
root directory. This new root directory needs an entry to point to
the “new” directory A, but directory B hasn't changed at
all. Therefore, our new root directory also has an entry that still
points to the <emphasis>old</emphasis> directory B node!</para>
<programlisting>
______________________________________________________
|___1_______2________3________4________5_________6_____...
|
|
___|_____ ________
|D | |D |
| | | |
| A | | A |
| \ | | \ |
| B \ | | B \ |
|__/___\__| |__/___\_|
/ \ / \
| ___\_____________/ \
| / \ \
___|__/ ___\____ ___\____
|D | |D | |D |
| | | | | |
| | | fish | | fish |
|_______| |___\____| |___\____|
\ \
\ \
___\____ ___\____
|D | |D |
| | | |
| tuna | | tuna |
|___\____| |___\____|
\ \
\ \
___\____ ___\____
|F | |F |
| | | |
| | | |
|________| |________|
</programlisting>
<para>Finally, after all our new nodes are written, we finish the
“bubble up” process by linking this new tree to the next
available revision in the history array. In this case, the new tree
becomes revision 2 in the repository.</para>
<programlisting>
______________________________________________________
|___1_______2________3________4________5_________6_____...
| \
| \__________
___|_____ __\_____
|D | |D |
| | | |
| A | | A |
| \ | | \ |
| B \ | | B \ |
|__/___\__| |__/___\_|
/ \ / \
| ___\_____________/ \
| / \ \
___|__/ ___\____ ___\____
|D | |D | |D |
| | | | | |
| | | fish | | fish |
|_______| |___\____| |___\____|
\ \
\ \
___\____ ___\____
|D | |D |
| | | |
| tuna | | tuna |
|___\____| |___\____|
\ \
\ \
___\____ ___\____
|F | |F |
| | | |
| | | |
|________| |________|
</programlisting>
<para>Generalizing on this example, you can now see that each
“revision” in the repository history represents a root
node of a unique tree (and an atomic commit to the whole filesystem.)
There are many trees in the repository, and many of them share
nodes.</para>
<para>Many nice behaviors come from this model:</para>
<orderedlist numeration="arabic">
<listitem><para><emphasis role="bold">Easy reads.</emphasis> If a
filesystem reader wants to locate revision
<replaceable>X</replaceable> of file <filename>foo.c</filename>,
it need only traverse the repository's history, locate revision
<replaceable>X</replaceable>'s root node, then walk down the tree
to <filename>foo.c</filename>.</para></listitem>
<listitem><para><emphasis role="bold">Writers don't interfere with
readers.</emphasis> Writers can continue to create new nodes,
bubbling their way up to the top, and concurrent readers cannot
see the work in progress. The new tree only becomes visible to
readers after the writer makes its final “link” to
the repository's history.</para></listitem>
<listitem><para><emphasis role="bold">File structure is
versioned.</emphasis> Unlike CVS, the very structure of each
tree is being saved from revision to revision. File and
directory renames, additions, and deletions are part of the
repository's history.</para></listitem>
</orderedlist>
<para>Let's demonstrate the last point by renaming the
<filename>tuna</filename> to <filename>book</filename>.</para>
<para>We start by creating a new parent “fish” directory,
except that this parent directory has a different dir_entry, one
which points the <emphasis>same</emphasis> old file node, but has a
different name:</para>
<programlisting>
______________________________________________________
|___1_______2________3________4________5_________6_____...
| \
| \__________
___|_____ __\_____
|D | |D |
| | | |
| A | | A |
| \ | | \ |
| B \ | | B \ |
|__/___\__| |__/___\_|
/ \ / \
| ___\_____________/ \
| / \ \
___|__/ ___\____ ___\____
|D | |D | |D |
| | | | | |
| | | fish | | fish |
|_______| |___\____| |___\____|
\ \
\ \
___\____ ___\____ ________
|D | |D | |D |
| | | | | |
| tuna | | tuna | | book |
|___\____| |___\____| |_/______|
\ \ /
\ \ /
___\____ ___\____ /
|F | |F |
| | | |
| | | |
|________| |________|
</programlisting>
<para>From here, we finish with the bubble-up process. We make new
parent directories up to the top, culminating in a new root directory
with two dir_entries (one points to the old “B” directory
node we've had all along, the other to the new revision of
“A”), and finally link the new tree to the history as
revision 3:</para>
<programlisting>
______________________________________________________
|___1_______2________3________4________5_________6_____...
| \ \_________________
| \__________ \
___|_____ __\_____ __\_____
|D | |D | |D |
| | | | | |
| A | | A | | A |
| \ | | \ | | \ |
| B \ | | B \ | | B \ |
|__/___\__| |__/___\_| |__/___\_|
/ ___________________/_____\_________/ \
| / ___\_____________/ \ \
| / / \ \ \
___|/_/ ___\____ ___\____ _____\__
|D | |D | |D | |D |
| | | | | | | |
| | | fish | | fish | | fish |
|_______| |___\____| |___\____| |___\____|
\ \ \
\ \ \
___\____ ___\____ ___\____
|D | |D | |D |
| | | | | |
| tuna | | tuna | | book |
|___\____| |___\____| |_/______|
\ \ /
\ \ /
___\____ ___\____ /
|F | |F |
| | | |
| | | |
|________| |________|
</programlisting>
<para>For our last example, we'll demonstrate the way
“tags” and “branches” are implemented in the
repository.</para>
<para>In a nutshell, they're one and the same thing. Because nodes are
so easily shared, we simply create a <emphasis>new</emphasis>
directory entry that points to an existing directory node. It's an
extremely cheap way of copying a tree; we call this new entry a
<firstterm>clone</firstterm>, or more colloquially, a “cheap
copy”.</para>
<para>Let's go back to our original tree, assuming that we're at
revision 6 to begin with:</para>
<programlisting>
______________________________________________________
...___6_______7________8________9________10_________11_____...
|
|
___|_____
|D |
| |
| A |
| \ |
| B \ |
|__/___\__|
/ \
| \
| \
___|___ ___\____
|D | |D |
| | | |
| | | fish |
|_______| |___\____|
\
\
___\____
|D |
| |
| tuna |
|___\____|
\
\
___\____
|F |
| |
| |
|________|
</programlisting>
<para>Let's “tag” directory A. To make the clone, we
create a new dir_entry <emphasis role="bold">T</emphasis> in our
root, pointing to A's node:</para>
<programlisting>
______________________________________________________
|___6_______7________8________9________10_________11_____...
| \
| \
___|_____ __\______
|D | |D |
| | | |
| A | | A |
| \ | | | |
| B \ | | B | T |
|__/___\__| |_/__|__|_|
/ \ / | |
| ___\__/ / /
| / \ / /
___|__/ ___\__/_ /
|D | |D |
| | | |
| | | fish |
|_______| |___\____|
\
\
___\____
|D |
| |
| tuna |
|___\____|
\
\
___\____
|F |
| |
| |
|________|
</programlisting>
<para>Now we're all set. In the future, the contents of directories A
and B may change quite a lot. However, assuming we never make any
changes to directory T, it will <emphasis>always</emphasis> point to
a particular pristine revision of directory A at some point in time.
Thus, T is a tag.</para>
<para>(In theory, we can use some kind of authorization system to
prevent anyone from writing to directory T. In practice, a well-laid
out repository should encourage “tag directories” to live
in one place, so that it's clear to all users that they're not meant
to change.)</para>
<para>However, if we <emphasis>do</emphasis> decide to allow commits in
directory T, and now our repository tree increments to revision 8,
then T becomes a branch. Specifically, it's a branch of directory A
which shares history with A up to a certain point, and then
“broke off” from the main line at revision 8.</para>
</sect3>
<sect3 id="server.fs.struct.diffy-storage">
<title>Diffy Storage</title>
<para>You may have been thinking, “Gee, this bubble up method
seems nice, but it sure wastes a lot of space. Every commit to the
repository creates an entire line of new directory
nodes!”</para>
<para>Like many other revision control systems, Subversion stores
changes as differences. It doesn't make complete copies of nodes;
instead, it stores the <emphasis>latest</emphasis> revision as a full
text, and previous revisions as a succession of reverse diffs (the
word "diff" is used loosely here – for files, it means vdeltas,
for directories, it means a format that expresses changes to
directories).</para>
</sect3>
</sect2>
<sect2 id="Implementation">
<title>Implementation</title>
<para>For the initial release of Subversion,</para>
<itemizedlist mark="bullet">
<listitem><para>The filesystem will be implemented as a library on
Unix.</para></listitem>
<listitem><para>The filesystem's data will probably be stored in a
collection of .db files, using the Berkeley Database library.
<footnote><para>In the future, ofcourse, contributors are free
modify the Subversion filesystem tooperate with more powerful
SQL database.</para></footnote> (For more information, see
<ulink url="http://www.sleepycat.com">Sleepycat
Software</ulink>.)</para></listitem>
</itemizedlist>
</sect2>
</sect1>
<sect1 id="server.libsvn_repos">
<title>Repository Library</title>
<!-- Jimb, Karl: Maybe we should turn this into a discussion about how the
filesystem will use non-historical properties for internal ACLs, and how
people can add "external" ACL systems via historical properties...? -->
<para>A subversion <firstterm>repository</firstterm> is a directory that
contains a number of components:</para>
<itemizedlist mark="bullet">
<listitem><para>a versioned filesystem (typically a collection of .db
files)</para></listitem>
<listitem><para>some hook scripts (for executing before or after
commits)</para></listitem>
<listitem><para>a locking area (used by Berkeley DB or other
processes)</para></listitem>
<listitem><para>a configuration area (for changing global
behaviors)</para></listitem>
</itemizedlist>
<para>The Subversion filesystem is just that: a filesystem. But it's also
useful to provide an API that acts at the level of the repository. The
repository library (<filename>libsvn_repos</filename>) does this.</para>
<para>In particular, it wraps a few <filename>libsvn_fs</filename>
routines, such as those for beginning and ending commits, so that
hook-scripts can run. A pre-commit-hook script might check for a valid
log message, and a post-commit-hook script might send an email to a
mailing list.</para>
<para>Additionally, the repository library provides convenience routines
for examining and manipulating the filesystem. For example, a routine to
generate a tree-delta by comparing two revisions, routines for
constructing new transactions, routines for querying log messages, and
routines for exporting and importing filesystem data.</para>
</sect1>
</chapter>