impl.html

<!DOCTYPE html>
<html>
<head>
<link rel="stylesheet" type="text/css" href="doc.css" />
<title>Leveldb file layout and compactions</title>
</head>

<body>

<h1>Files</h1>

The implementation of leveldb is similar in spirit to the
representation of a single
<a href="http://research.google.com/archive/bigtable.html">
Bigtable tablet (section 5.3)</a>.
However the organization of the files that make up the representation
is somewhat different and is explained below.

<p>
Each database is represented by a set of files stored in a directory.
There are several different types of files as documented below:
<p>
<h2>Log files</h2>
<p>
A log file (*.log) stores a sequence of recent updates.  Each update
is appended to the current log file.  When the log file reaches a
pre-determined size (approximately 4MB by default), it is converted
to a sorted table (see below) and a new log file is created for future
updates.
<p>
A copy of the current log file is kept in an in-memory structure (the
<code>memtable</code>).  This copy is consulted on every read so that read
operations reflect all logged updates.
<p>
<h2>Sorted tables</h2>
<p>
A sorted table (*.sst) stores a sequence of entries sorted by key.
Each entry is either a value for the key, or a deletion marker for the
key.  (Deletion markers are kept around to hide obsolete values
present in older sorted tables).
<p>
The set of sorted tables are organized into a sequence of levels.  The
sorted table generated from a log file is placed in a special <code>young</code>
level (also called level-0).  When the number of young files exceeds a
certain threshold (currently four), all of the young files are merged
together with all of the overlapping level-1 files to produce a
sequence of new level-1 files (we create a new level-1 file for every
2MB of data.)
<p>
Files in the young level may contain overlapping keys.  However files
in other levels have distinct non-overlapping key ranges.  Consider
level number L where L >= 1.  When the combined size of files in
level-L exceeds (10^L) MB (i.e., 10MB for level-1, 100MB for level-2,
...), one file in level-L, and all of the overlapping files in
level-(L+1) are merged to form a set of new files for level-(L+1).
These merges have the effect of gradually migrating new updates from
the young level to the largest level using only bulk reads and writes
(i.e., minimizing expensive seeks).

<h2>Manifest</h2>
<p>
A MANIFEST file lists the set of sorted tables that make up each
level, the corresponding key ranges, and other important metadata.
A new MANIFEST file (with a new number embedded in the file name)
is created whenever the database is reopened.  The MANIFEST file is
formatted as a log, and changes made to the serving state (as files
are added or removed) are appended to this log.
<p>
<h2>Current</h2>
<p>
CURRENT is a simple text file that contains the name of the latest
MANIFEST file.
<p>
<h2>Info logs</h2>
<p>
Informational messages are printed to files named LOG and LOG.old.
<p>
<h2>Others</h2>
<p>
Other files used for miscellaneous purposes may also be present
(LOCK, *.dbtmp).

<h1>Level 0</h1>
When the log file grows above a certain size (1MB by default):
<ul>
<li>Create a brand new memtable and log file and direct future updates here
<li>In the background:
<ul>
<li>Write the contents of the previous memtable to an sstable
<li>Discard the memtable
<li>Delete the old log file and the old memtable
<li>Add the new sstable to the young (level-0) level.
</ul>
</ul>

<h1>Compactions</h1>

<p>
When the size of level L exceeds its limit, we compact it in a
background thread.  The compaction picks a file from level L and all
overlapping files from the next level L+1.  Note that if a level-L
file overlaps only part of a level-(L+1) file, the entire file at
level-(L+1) is used as an input to the compaction and will be
discarded after the compaction.  Aside: because level-0 is special
(files in it may overlap each other), we treat compactions from
level-0 to level-1 specially: a level-0 compaction may pick more than
one level-0 file in case some of these files overlap each other.

<p>
A compaction merges the contents of the picked files to produce a
sequence of level-(L+1) files.  We switch to producing a new
level-(L+1) file after the current output file has reached the target
file size (2MB).  We also switch to a new output file when the key
range of the current output file has grown enough to overlap more then
ten level-(L+2) files.  This last rule ensures that a later compaction
of a level-(L+1) file will not pick up too much data from level-(L+2).

<p>
The old files are discarded and the new files are added to the serving
state.

<p>
Compactions for a particular level rotate through the key space.  In
more detail, for each level L, we remember the ending key of the last
compaction at level L.  The next compaction for level L will pick the
first file that starts after this key (wrapping around to the
beginning of the key space if there is no such file).

<p>
Compactions drop overwritten values.  They also drop deletion markers
if there are no higher numbered levels that contain a file whose range
overlaps the current key.

<h2>Timing</h2>

Level-0 compactions will read up to four 1MB files from level-0, and
at worst all the level-1 files (10MB).  I.e., we will read 14MB and
write 14MB.

<p>
Other than the special level-0 compactions, we will pick one 2MB file
from level L.  In the worst case, this will overlap ~ 12 files from
level L+1 (10 because level-(L+1) is ten times the size of level-L,
and another two at the boundaries since the file ranges at level-L
will usually not be aligned with the file ranges at level-L+1).  The
compaction will therefore read 26MB and write 26MB.  Assuming a disk
IO rate of 100MB/s (ballpark range for modern drives), the worst
compaction cost will be approximately 0.5 second.

<p>
If we throttle the background writing to something small, say 10% of
the full 100MB/s speed, a compaction may take up to 5 seconds.  If the
user is writing at 10MB/s, we might build up lots of level-0 files
(~50 to hold the 5*10MB).  This may signficantly increase the cost of
reads due to the overhead of merging more files together on every
read.

<p>
Solution 1: To reduce this problem, we might want to increase the log
switching threshold when the number of level-0 files is large.  Though
the downside is that the larger this threshold, the more memory we will
need to hold the corresponding memtable.

<p>
Solution 2: We might want to decrease write rate artificially when the
number of level-0 files goes up.

<p>
Solution 3: We work on reducing the cost of very wide merges.
Perhaps most of the level-0 files will have their blocks sitting
uncompressed in the cache and we will only need to worry about the
O(N) complexity in the merging iterator.

<h2>Number of files</h2>

Instead of always making 2MB files, we could make larger files for
larger levels to reduce the total file count, though at the expense of
more bursty compactions.  Alternatively, we could shard the set of
files into multiple directories.

<p>
An experiment on an <code>ext3</code> filesystem on Feb 04, 2011 shows
the following timings to do 100K file opens in directories with
varying number of files:
<table class="datatable">
<tr><th>Files in directory</th><th>Microseconds to open a file</th></tr>
<tr><td>1000</td><td>9</td>
<tr><td>10000</td><td>10</td>
<tr><td>100000</td><td>16</td>
</table>
So maybe even the sharding is not necessary on modern filesystems?

<h1>Recovery</h1>

<ul>
<li> Read CURRENT to find name of the latest committed MANIFEST
<li> Read the named MANIFEST file
<li> Clean up stale files
<li> We could open all sstables here, but it is probably better to be lazy...
<li> Convert log chunk to a new level-0 sstable
<li> Start directing new writes to a new log file with recovered sequence#
</ul>

<h1>Garbage collection of files</h1>

<code>DeleteObsoleteFiles()</code> is called at the end of every
compaction and at the end of recovery.  It finds the names of all
files in the database.  It deletes all log files that are not the
current log file.  It deletes all table files that are not referenced
from some level and are not the output of an active compaction.

</body>
</html>