MDB: A Memory-Mapped Database and Backend for OpenLDAP

Howard Chu
Symas Corp., OpenLDAP Project
[email protected], [email protected]
http://www.symas.com, http://www.openldap.org

Abstract
This paper introduces MDB ("Memory-Mapped Database"), a read-optimized database library and slapd
backend developed for OpenLDAP. In this paper we will discuss OpenLDAP's traditional primary database as
well as some other alternatives that were examined before arriving at the MDB implementation. Early results
from testing the new MDB implementation will also be presented. This work is in progress but essentially
complete, and will be integrated into the OpenLDAP public releases in the near future.

1. Introduction
While OpenLDAP already provides a reliable high performance transactional backend
database (using Oracle BerkeleyDB "BDB"[1]), it requires careful tuning to get good results
and the tuning aspects can be quite complex. Data comes through three separate layers of
caches before it may be used, and each cache layer has a significant footprint. Balancing the
three layers against each other can be a difficult juggling act. Additionally, there are two layers
of locking needed to safely manipulate these caches, and the locks severely limit the
scalability of the database on multi-processor machines.
Rather than continue to attempt to adapt other third-party database software into OpenLDAP,
the MDB library was written specifically for use in OpenLDAP. The library is fully transactional
and implements B+ trees[2] with Multi-Version Concurrency Control[3]. The entire database is
mapped into virtual memory and all data fetches are performed via direct access to the
mapped memory instead of through intermediate buffers and copies.

2. Background
Before describing the improvements offered by the MDB design, an overview of the existing
BDB-based backends (back-bdb and back-hdb) will be presented.
LDAP and BDB have a long history together; Netscape commissioned the 2.0 release of BDB
specifically for use in their LDAP server[4]. The OpenLDAP Project's first release using the
BDB-specific APIs was OpenLDAP 2.1 in June 2002. Since BDB maintains its own internal
cache, it was hoped that the back-bdb backend could be deployed without any backend-level
caching, but early benchmark results showed that retrieving entries directly from the database
on every query was too slow. Despite radical improvements in entry fetch and decoding
speed[5], the decision was made to introduce an entry cache for the backend, and the cache
management problems grew from there.
These problems include:
• Multiple caches that each need to be carefully configured. On top of the BDB cache,
there are caches for entries, DNs, and attribute indexing in the backend. All of these
waste memory since the same data may be present in three places - the filesystem
cache, the BDB cache, and the backend caches. Configuration is a tedious job
because each cache layer has different size and speed characteristics and it is difficult
to strike a balance that is optimal for all use cases.
• Caches with very complex lock dependencies. For speed, most of the backend caches
are protected by simple mutexes. However, when interacting with the BDB API, these
mutexes must be dropped and exchanged for (much slower) database locks.
Otherwise deadlocks which could not be detected by BDB's deadlock detector may
occur. Deadlocks occur very frequently in routine operation of the backend.
• Caches with pathological behavior if they were smaller than the whole database. When
the cache size was small enough that a significant number of queries were not being
satisfied from the cache, extreme heap fragmentation was observed[6], as the cache
freed existing entries to make room for new entries. The fragmentation would cause
the size of the slapd process to rapidly grow, defeating the purpose of setting a small
cache size. The problem was worst with the memory allocator in GNU libc[7], and
could be mitigated by using alternatives such as Hoard[8] or Google tcmalloc[9], but
additional changes were made in slapd to reduce the number of calls to malloc() and
free() to delay the onset of this fragmentation issue[10].
• Caches with very low effectiveness. When multiple queries arrive whose result sets are
larger than the entry cache, the cache effectiveness drops to zero because entries are
constantly being freed before they ever get any chance of being re-used[11]. A great
deal of effort was expended exploring more advanced cache replacement algorithms to
combat this problem[12][13].
From the advent of the back-bdb backend until the present time, the majority of development
and debugging effort in these backends has all been devoted to backend cache management.
The present state of affairs is difficult to configure, difficult to optimize, and extremely labor
intensive to maintain.
Another issue relates to administrative overhead in general. For example, BDB uses write-
ahead logs for its transaction support. These logs are written before database updates are
performed, so that in case an update is interrupted or aborted, sufficient information is present
to undo the updates and return the database to the state it was in before the update began.
The log files grow continuously as updates are made to a database, and can only be removed
after an expensive checkpoint operation is performed. Later versions of BDB added an auto-
remove option to delete old log files automatically, but if the system crashed while this option
was in use, generally the database could not be recovered successfully because the
necessary logs had been deleted.
3. Solutions
The problems with back-bdb and back-hdb can be summed up in two main areas: cache
management, and lock management. The approach to a solution with back-mdb is simple - do
no caching, and do no locking. The other issues of administrative overhead are handled as
side-effects of the main solutions.

3.1 Eliminating Caching

One fundamental concept behind the MDB approach is known as "Single-Level Store"[14].
The basic idea is to treat all of computer memory as a single address space. Pages of
storage may reside in primary storage (RAM) or in secondary storage (disk) but the actual
location is unimportant to the application. If a referenced page is currently in primary storage
the application can use it immediately, if not a page fault occurs and the operating system
brings the page into primary storage. The concept was introduced in 1964 in the Multics[15]
operating system but was generally abandoned by the early 1990s as data volumes
surpassed the capacity of 32 bit address spaces. (We last knew of it in the Apollo
DOMAIN[16] operating system, though many other Multics-influenced designs carried it on.)
With the ubiquity of 64 bit processors today this concept can again be put to good use. (Given
a virtual address space limit of 63 bits that puts the upper bound of database size at 8
exabytes. Commonly available processors today only implement 48 bit address spaces,
limiting us to 47 bits or 128 terabytes.)
Another operating system requirement for this approach to be viable is a Unified Buffer
Cache. While most POSIX-based operating systems have supported an mmap() system call
for many years, their initial implementations kept memory managed by the VM subsystem
separate from memory managed by the filesystem cache. This was not only wasteful (again,
keeping data cached in two places at once) but also led to coherency problems - data
modified through a memory map was not visible using filesystem read() calls, or data modified
through a filesystem write() was not visible in the memory map. Most modern operating
systems now have filesystem and VM paging unified, so this should not be a concern in most
deployments[17][18][19].
The MDB library is designed to access an entire database thru a single read-only memory
map. Keeping the mapping read-only prevents stray writes from buggy application code from
corrupting the database. Updates are performed using regular write() calls. (Updating through
the map would be difficult anyway since files cannot be grown through map references; only
updates to existing pages could be done through the map. For simplicity all updates are done
using write() and it doesn't matter whether the update grows the file or not.) This update
approach requires that the filesystem and VM views are kept coherent, thus the requirement
that the OS uses a Unified Buffer Cache.
The memory-mapped approach makes full use of the operating system's filesystem cache,
and eliminates any database-level caching. Likewise the back-mdb backend performs no
caching of its own; it uses information from the database directly. Using the memory-mapped
data thus eliminates two levels of caching relative to back-hdb, as well as eliminating
redundant memcpy() operations between those caches. It also eliminates all cache
tuning/configuration issues, thus easing deployment for administrators.
Of course, by eliminating caching, one would expect to incur a significant performance hit. It
should be much faster to dump out the contents of a cached, fully decoded entry in response
to a search request, than to read the entry in from disk and decode it on every request. Early
results with back-mdb showed this to be true, but further optimization in back-mdb has mostly
eliminated this performance hit.

3.2 Eliminating Locking

The other fundamental concept behind MDB is the use of Multi-Version Concurrency Control
(MVCC). The basic idea is that updates of data never overwrite existing data; instead the
updates write to new pages and thus create a new version of the database. Readers only
ever see the snapshot of the database as it existed when a read transaction began, so they
are fully isolated from writers. Because of this isolation read accesses require no locks, they
always have a self-consistent view of the database.
BDB has supported MVCC since version 4.5.20, but because of the caching layer in back-
bdb/hdb there was no benefit to using it. The only way to get any gain from using MVCC was
to also eliminate the backend caching layer, and without the caching layer back-bdb/hdb's
performance would be too slow because data lookups in BDB were still too slow.
A major downside of MVCC-based systems is that since they always write new data to new
disk pages, the database files tend to grow without bound. They need periodic compaction or
garbage collection in order to keep their disk usage constrained, and the required frequency
of such compaction efforts is very high on databases with high update rates. Additionally,
systems based on garbage collection generally require twice as much disk space as the
actual data occupies. Also, in order to sustain a write rate of N operations/second, the I/O
system must actually support >>2N operations/second, since the compaction task needs to
run faster than the normal write task in order to catch up and actually complete its job, and the
volume of data already written always exceeds the volume being written. If this over-
provisioning of I/O resources cannot be guaranteed, then the typical solution to this problem
is to deny updates while compaction is being performed.
Causing a service outage for writes while garbage collection is performed is unacceptable, so
MDB uses a different approach. Within a given MDB database environment, MDB maintains
two B+tree structures - one containing application data, and another one containing a free list
with the IDs of pages that are no longer in use. Tracking the in-use status is typically done
with reference counters and other such mechanisms that require locking. Obviously the use of
locking would defeat the purpose of using MVCC in the first place, so a lockless solution was
designed instead. With this solution, pages that are no longer in use by any active snapshot
of the database are re-used by updaters, so the database size remains relatively static. This
is a key advantage of MDB over other well-known MVCC databases such as CouchDB[20].
4. Implementation Highlights
Since all of the source code has been publicly available from the outset, and due to space
limitations in this paper, only a few of the most notable implementation details will be
described here. Interested parties are invited to read the code in the OpenLDAP git repository
and post questions on the openldap-technical mailing list.
The MDB library API was loosely modeled after the BDB API, to ease migration of BDB-based
code. The first cut of the back-mdb code was simply copied from the back-bdb source tree,
and then all references to the caching layers were deleted. After a few minor API differences
were accounted for, the backend was fully operational (though still in need of optimization). As
of today back-mdb comprises 340KB of source code, compared to 476KB for back-bdb/hdb,
so back-mdb is approximately 30% smaller.
The MDB code itself started from Martin Hedenfalk's append-only Btree code in the OpenBSD
ldapd source repository[21]. The first cut of the MDB code was simply copied from the ldapd
source, and then all of the Btree page cache manager was deleted and replaced with mmap
accesses. The original Btree source yielded an object file of 39KB; the MDB version was
32KB. Initial testing with the append-only code proved that approach to be completely
impractical. With a small test database and only a few hundred add/delete operations, the DB
occupied 1027 pages but only 10 pages actually contained current data; over 99% of the
space was wasted.
Along with the mmap management and page reclamation, many other significant changes
were made to arrive at the current MDB library, mostly to add features from BDB that back-
mdb would need. As of today the MDB library comprises 35KB of object code. (Comparing
source code is not very informative since the MDB source code has been heavily expanded
with Doxygen comments. The initial version of mdb.c was 59KB as opposed to btree.c at
76KB but with full documentation embedded mdb.c is now 162KB. Also for comparison, BDB
is now over 1.5MB of object code.)

4.1 MDB Change Summary

The append-only Btree code used a meta page at the end of the database file to point at the
current root node of the Btree. New pages were always written out sequentially at the end of
the file, followed by a new meta page upon transaction commit. Any application opening the
database needed to search backward from the end of the file to find the most recent meta
page, to get a current snapshot of the database. (Further explanation of append-only
operation is available at Martin's web site[22].)
In MDB there are two meta pages occupying page 0 and page 1 of the file. They are used
alternately by transactions. Each meta page points to the root node of two Btrees - one for the
free list and one for the application data. New data first re-uses any available pages from the
free list, then writes sequentially at the end of the file if no free pages are available. Then the
older meta page is written on transaction commit. This is nothing more than standard double-
buffering - any application opening the database uses the newer meta page, while a
committer overwrites the older one. No locks are needed to protect readers from writers;
readers are guaranteed to always see a valid root node.
The original code only supported a single Btree in a given database file. For MDB we wanted
to support multiple trees in a single database file. The back-mdb indexing code uses
individual databases for each attribute index, and it would be a non-starter to require a
sysadmin to configure multiple mmap regions for a single back-mdb instance. Additionally, the
indexing code uses BDB's sorted duplicate feature, which allows multiple data items with the
same key to be stored in a Btree, and this feature needed to be added to MDB as well. These
features were both added using a subdatabase mechanism, which allows a data item in a
Btree to be treated as the root node of another Btree.

4.2 Locking
For simplicity the MDB library allows only one writer at a time. Creating a write transaction
acquires a lock on a writer mutex; the mutex normally resides in a shared memory region so
that it can be shared between multiple processes. This shared memory is separate from the
region occupied by the main database. The lock region also contains a table with one slot for
every active reader in the database. The slots record the reader's process and thread ID, as
well as the ID of the transaction snapshot the reader is using. (The process and thread ID are
recorded to allow detection of stale entries in the table, e.g. threads that exited without
releasing their reader slot.) The table is constructed in processor cache-aligned memory such
that False Sharing[23] of cache lines is avoided.
Readers acquire a slot the first time a thread opens a read transaction. Acquiring an empty
slot in the table requires locking a mutex on the table. The slot address is saved in thread-
local storage and re-used the next time the thread opens a read transaction, so the thread
never needs to touch the table mutex ever again. The reader stores its transaction ID in the
slot at the start of the read transaction and zeroes the ID in the slot at the end of the
transaction. In normal operation, there is nothing that can block the operation of readers.
The reader table is used when a writer wants to allocate a page, and knows that the free list is
not empty. Writes are performed using copy-on-write semantics; whenever a page is to be
written, a copy is made and the copy is modified instead of the original. Once copied, the
original page's ID is added to an in-memory free list. When a transaction is committed, the in-
memory free list is saved as a single record in the free list DB along with the ID of the
transaction for this commit. When a writer wants to pull a page from the free list DB, it
compares the transaction ID of the oldest record in the free list DB with the transaction IDs of
all of the active readers. If the record in the free list DB is older than all of the readers, then
all of the pages in that record may be safely re-used because nothing else in the DB points to
them any more.
The writer's scan of the reader table also requires no locks, so readers cannot block writers.
The only consequence of a reader holding onto an old snapshot for a long time is that page
reclaiming cannot be done; the writer will simply use newly allocated pages in the meantime.
4.3 Backend Features
The database layout in back-mdb is functionally identical to the one used in back-hdb so it is
also fully hierarchical. Entries are stored in a binary format based on the one used for back-
hdb, but with further encoding optimizations. The most significant optimization was to use a
mapping of AttributeDescriptions to small integers, so that their canonical names were no
longer stored in each entry. This saved a bit of space in the encoded entry, but more
importantly made Attribute decoding an O(1) operation instead of O(logN). Also, while the
MDB library doesn't need to allocate any memory to return data, entries still require Entry and
Attribute structures to be allocated. But since entries don't need to be kept persistently in a
cache, all allocations can be done from temporary thread-local memory. As a result of these
optimizations the entry decoder is 6.5 times faster overall than the one used in back-hdb.
Configuration for back-mdb is much simplified - there are no cache configuration directives.
The backend requires only a pathname for storing the database files, and a maximum allowed
size for the database. The configuration settings only affect the capacity of the database, not
its performance; there is nothing to tune.

5. Results
Profiling was done using multiple tools, including FunctionCheck[24], valgrind callgrind[25],
and oprofile[26], to aid in optimization of MDB. Oprofile has the least runtime overhead and
provides the best view of multi-threaded behavior, but since it is based on random samples it
tends to miss some data of interest. FunctionCheck is slower, at four times slower than
normal, but since it uses instrumented code it always provides a complete profile of overall
function run times. callgrind is slowest, at thirty times slower than normal, and only provides
relevant data for single-threaded operation, but since it does instruction-level profiling it gives
the most detailed view of program behavior. Since program behavior can vary wildly between
single-threaded and multi-processor operation, it was important to gather performance data
from a number of different perspectives.
Table 1 compares basic performance of back-mdb vs back-hdb for initially loading a test
database using slapadd in "quick" mode.

real user sys

back-hdb 66m09.831s 115m52.374s 5m15.860s
back-mdb 29m33.212s 22m21.264s 7m11.851s
Table 1: Time to slapadd -q 5 million entries
back-hdb has a much higher user time than real time because it was using multi-threaded
indexing. At present back-mdb doesn't support multi-threaded operation for slapadd. back-
hdb was using BDB 4.7.25 in these tests, but results with BDB 5.2.28 were essentially the
same.
With the databases loaded, the next test was to start up slapd and time how long it took to
scan the entire database with a single ldapsearch. Also the slapd process sizes were
compared, relative to their DB sizes on disk. These results are summarized in Table 2.

first second slapd size DB size

back-hdb 4m15.395s 0m16.204s 26GB 15.6GB
back-mdb 0m14.725s 0m10.807s 10GB 12.8GB
Table 2: ldapsearch comparison
back-hdb is configured with an entry cache size of 5 million entries, so all of the database is
fully cached after the first ldapsearch is run. Also note that the DB files are entirely resident in
the filesystem cache since slapadd had just completed before. Also the BDB cache was
configured at 32GB so the entire database is resident there too; no disk I/O occurs during
these tests. This table shows the overhead of retrieving data from the BDB cache and
decoding it into the back-hdb entry cache. But even with that overhead eliminated in the
second run, back-mdb is still faster. For back-mdb the extra time required in the first run
reflects the time needed for the OS to map the database pages into the slapd process'
address space. The slapd process size for mdb is smaller than the DB size for a couple of
reasons: first, the DB contains attribute indices, and this search doesn't reference any indices,
so those pages are not mapped into the process. second, the DB contains a number of free
pages that were left over from the last slapadd transaction.
Before development began it was estimated that the MDB approach would use 1/3 to 1/2 as
much RAM as the equivalent back-hdb database; this estimate was confirmed with back-mdb
using only 37% as much RAM as back-hdb on our fully cached test database.
Next, a basic concurrency test was performed by running the same ldapsearch operation 2, 4,
8, and 16 times concurrently and measuring the time to obtain the results. The averages of
the result times are shown in Table 3.

2 4 8 16
back-hdb, debian 0m23.147s 0m30.384s 1m25.665s 17m15.114s
back-hdb 0m24.617s 0m32.171s 1m04.817s 3m04.464s
back-mdb 0m10.789s 0m10.842s 0m10.931s 0m12.023s
Table 3: Concurrent Search Times
The first time this test was run with back-hdb yielded some extraordinarily poor results. Later
testing revealed that this test was accidentally run using the stock build of BDB 4.7 provided
by Debian, instead of the self-compiled build we usually use in our testing. The principle
difference is that we always build BDB with the configure option --with-mutex=POSIX/pthread,
whereas by default BDB uses a hybrid of spinlocks and pthread mutexes. The spinlocks are
fairly efficient within a single CPU socket, but they scale extremely poorly as the number of
processors increases. back-mdb's scaling is essentially flat across arbitrary numbers of
processors since it has no locking to slow it down. The performance degrades slightly at the
16 search case because at that point all of the processors on our test machine are busy so
the clients and slapd are competing with other system processes for CPU time. As another
point of reference, the time required to copy the MDB database to /dev/null using 'dd' was
10.20 seconds. Even with all of the decoding and filtering that slapd needed to do, scanning
the entire DB was only 6% slower than a raw copy operation.
The previous tests show worst-case performance for search operations. For more real-world
results, we move on to using SLAMD[27]. (SLAMD has known performance issues, but we've
gotten used to them, and staying with the same tool lets us compare with historical results
from our previous work as well.) Table 4 summarizes the results for back-hdb vs back-mdb
with randomly generated queries across the 5 million entry database.

Searches/sec Duration, msec

back-hdb 67456.11 1.89
back-mdb 119255.42 0.63
Table 4: SLAMD Search Rate Results
The back-hdb result is actually extremely good - it's about 15% faster than the second fastest
directory software we've tested previously on this machine (OpenDS 2.3). But they're all
utterly outclassed by back-mdb. If you look at the actual stats in Illustration 1 you'll see that
the performance was still increasing as the process' page map was filling in.

Illustration 1: back-mdb Search Rate results

After seeing these results we considered renaming MDB as "LightningDB" - its read
performance is blindingly fast and totally unparalleled.
For write speeds, back-mdb is significantly slower than back-hdb. Table 5 shows the
throughput in a pure Modify test, modifying a single attribute in random entries across the 5
million entry database.

Modifies/sec Duration, msec

back-hdb 20440.83 1.56
back-mdb 6131.77 1.29
Table 5: SLAMD Modify Rate Results
Note that back-mdb actually completes modifies quickly, but because MDB enforces single-
writer behavior, it does not accept as many writes per second. Our final comparison in Table 6
shows a Modify Rate job running concurrently with a Search Rate job.

Searches/sec Search msec Modifies/sec Modify msec

back-hdb 40629.49 1.47 12321.36 1.62
back-mdb 85918.92 1.77 2844.95 2.80
Table 6: SLAMD Combined Search and Modify Rate
Most of the effort has been focused on read performance so far; future work may
be able to boost MDB's write performance but it is not perceived as a critical
problem for now.

6. Conclusions
The combination of memory-mapped operation with Multi-Version Concurrency
Control proves to be extremely potent for LDAP directories. The administrative
overhead is minimal since MDB databases require no periodic cleanup or garbage
collection, and no particular tuning is needed. Code size and complexity have
been drastically reduced, while read performance has been significantly raised.
Write performance has been traded for read performance, but this is acceptable
and can be addressed in more depth in the future.

6.1 Portability
While initial development was done on Linux, MDB and back-mdb have been
ported to MacOSX and Windows. No special problems are anticipated in porting to
other platforms.

6.2 Other Directions

A port of SQLite to use MDB has also been done. The MDB library needed to be
extended to support nested transactions, but otherwise needed very little
changes. Basic functionality is working already, and the code can be obtained at
http://gitorious.org/mdb/sqlightning. There are probably many other applications
for a small-footprint database library with relatively low write rates and near-zero
read overhead.

6.3 Future Work

A number of items remain on our ToDo list.
• Write optimization has not yet been investigated.
• A bulk-loading API for further speedups in slapadd would be nice.
• It would be nice to allow the database map size and other configuration
settings to be grown dynamically instead of statically configured.
• Functions to facilitate incremental and/or full backups would be nice to
have.
• A back-mdb that stores entries in the DB in their in-memory format, thus
requiring no decoding at all, is still being considered.
None of these items are seen as critical show-stoppers. MDB and back-mdb
already meet all the goals set for them and fulfill all of the functions required of an
OpenLDAP backend, while setting a new standard for database efficiency,
scalability, and performance.

References
1: Oracle, BerkeleyDB, 2011,
http://www.oracle.com/technetwork/database/berkeleydb/overview/index.html
2: Wikipedia, B+trees, , http://en.wikipedia.org/wiki/B+_tree
3: Wikipedia, MVCC, , http://en.wikipedia.org/wiki/Multiversion_concurrency_control
4: Margo Seltzer, Keith Bostic, Berkeley DB, The Architecture of Open Source Applications, 2011,
http://www.aosabook.org/en/bdb.html
5: Jong-Hyuk Choi and Howard Chu, Dissection of Search Latency, 2001,
http://www.openldap.org/lists/openldap-devel/200111/msg00042.html
6: Howard Chu, Better malloc strategies?, 2006, http://www.openldap.org/lists/openldap-
devel/200607/msg00005.html
7: Howard Chu, Malloc Benchmarking, 2006, http://highlandsun.com/hyc/malloc/
8: Emery Berger, The Hoard Memory Allocator, 2006-2010, http://www.hoard.org
9: Sanjay Ghemawat, Paul Menage, TCMalloc: Thread-Caching Malloc, 2005, http://goog-
perftools.sourceforge.net/doc/tcmalloc.html
10: Howard Chu, Minimize malloc fragmentation, 2006, http://www.openldap.org/lists/openldap-
devel/200608/msg00033.html
11: Wikipedia, Page replacement algorithms, ,
http://en.wikipedia.org/wiki/Page_replacement_algorithm#Least_recently_used
12: Howard Chu, CLOCK-Pro cache replacement code, 2007, http://www.openldap.org/lists/openldap-
bugs/200701/msg00039.html
13: Howard Chu, Cache-thrashing protection, 2007, http://www.openldap.org/lists/openldap-
commit/200711/msg00068.html
14: Wikipedia, Single-Level Store, , http://en.wikipedia.org/wiki/Single-level_store
15: Multicians, Multics General Information and FAQ, , http://www.multicians.org/general.html
16: Apollo Computer Inc., Domain/OS Design Principles, 1989, http://bitsavers.org/pdf/apollo/014962-
A00_Domain_OS_Design_Principles_Jan89.pdf
17: R.A. Gingell, J.P. Moran, and W.A. Shannon, Virtual Memory Architecture in SunOS, USENIX
Summer Conference, 1987, http://citeseer.ist.psu.edu/viewdoc/summary?doi=10.1.1.132.8931
18: Marshall Kirk McKusick, Keith Bostic, Michael J. Karels, John S. Quarterman, Memory
Management, The Design and Implementation of the 4.4BSD Operating System, 1996,
http://www.freebsd.org/doc/en/books/design-44bsd/overview-memory-management.html
19: Linus Torvalds, Status of the buffer cache in 2.3.7+, 1999,
http://lkml.indiana.edu/hypermail/linux/kernel/9906.3/0670.html
20: Apache Software Foundation, Apache CouchDB: Technical Overview, 2008-2011,
http://couchdb.apache.org/docs/overview.html
21: Martin Hedenfalk, OpenBSD ldapd source repository, 2010-2011, http://www.openbsd.org/cgi-
bin/cvsweb/src/usr.sbin/ldapd/
22: Martin Hedenfalk, How the Append-Only Btree Works, 2011, http://www.bzero.se/ldapd/btree.html
23: Suntorn Sae-eung, Analysis of False Cache Line Sharing Effects on Multicore CPUs, 2010,
http://scholarworks.sjsu.edu/etd_projects/2
24: Howard Chu, FunctionCheck, 2005, http://highlandsun.com/hyc/#fncchk
25: Valgrind Developers, Callgrind: a call-graph generating cache and branch prediction profiler, 2011,
http://valgrind.org/docs/manual/cl-manual.html
26: , OProfile - A System Profiler for Linux, 2011, http://oprofile.sourceforge.net/news/
27: UnboundID Corp., SLAMD Distributed Load Generation Engine, 2010, http://www.slamd.com