1 \documentclass[10pt]{article}
2 \usepackage[ansinew]{inputenc}
3 \usepackage{makeidx,amsmath,amssymb,exscale,multicol,epsfig,graphics,verbatim,ulem}
4 \usepackage{epsfig,geometry,url,listings, subcaption}
5 \usepackage{boxedminipage}
6 \usepackage[T1]{fontenc}%required
8 \geometry{headsep=3ex,hscale=0.9}
10 \hypersetup{pdftitle={GNUnet C Tutorial},
12 pdfauthor={Christian Grothoff <christian@grothoff.org>},
13 pdfkeywords={p2p,search,gnunet,tutorial}
14 %,pdfpagemode={FullScreen}
23 literate={*}{{\char42}}1
27 \newcommand{\exercise}[1]{\noindent\begin{boxedminipage}{\textwidth}{\bf Exercise:} #1 \end{boxedminipage}}
32 \large {A Tutorial for GNUnet 0.9.x (C version)}
34 Christian Grothoff $\qquad$ Bart Polot $\qquad$ Matthias Wachs
38 This tutorials explains how to install GNUnet on a GNU/Linux system ond gives an introduction how
39 GNUnet can be used to develop a Peer-to-Peer application. Detailed installation instructions for
40 various operating systems and a detailed list of all dependencies can found on our website at
41 \url{https://gnunet.org/installation}.
43 \textbf{Please read this tutorial carefully since every single step is important and do not hesitate to contact the GNUnet team if you have any questions or problems! Check here how to contact the GNUnet team:
44 \url{https://gnunet.org/contact_information}}
47 \section{Installing GNUnet}
48 First of all you have to install a current version of GNUnet. You can download a
49 tarball of a stable version from GNU FTP mirrors or obtain the latest development
50 version from our Subversion repository.
52 Most of the time you should prefer to download the stable version since with the
53 latest development version things can be broken, functionality can be changed or tests
54 can fail. You should only use the development version if you know that you require a
55 certain feature or a certain issue has been fixed since the last release.
57 \subsection{Obtaining a stable version}
58 You can download the latest stable version of GNUnet from GNU FTP mirrors:
60 \url{ftp://ftp.gnu.org/gnu/gnunet/gnunet-0.9.5a.tar.gz}
62 You should also download the signature file and verify the integrity of the tarball.
64 \url{ftp://ftp.gnu.org/gnu/gnunet/gnunet-0.9.5a.tar.gz.sig}
66 To verify the signature you should first import the GPG key used to sign the tarball
68 $ gpg --keyserver keys.gnupg.net --recv-keys 48426C7E
70 And use this key to verify the tarball's signature
72 $ gpg --verify gnunet-0.9.5a.tar.gz.sig gnunet-0.9.5a.tar.gz
74 After successfully verifying the integrity you can extract the tarball using
76 $ tar xvzf gnunet-0.9.5a.tar.gz
77 $ mv gnunet-0.9.5a gnunet # we will use the directory "gnunet" in the reminder of this document
81 \subsection{Installing Build Tool Chain and Dependencies}
82 To successfully compile GNUnet you need the tools to build GNUnet and the required dependencies.
83 Please have a look at \url{https://gnunet.org/dependencies} for a list of required dependencies
84 and \url{https://gnunet.org/generic_installation} for specific instructions for your operating system.
86 Please check the notes at the end of the configure process about required dependencies.
88 For GNUNet bootstrapping support and the http(s) plugin you should install \texttt{libcurl}.
89 For the filesharing service you should install at least one of the datastore backends \texttt{mysql},
90 \texttt{sqlite} or \texttt{postgresql}.
92 \subsection{Obtaining the latest version from Subversion}
93 The latest development version can obtained from our Subversion (\textit{svn}) repository. To obtain
94 the code you need Subversion installed and checkout the repository using:
95 \lstset{language=bash}
97 $ svn checkout https://gnunet.org/svn/gnunet
99 After cloning the repository you have to execute
100 \lstset{language=bash}
107 \subsection{Compiling and Installing GNUnet}
109 Assuming all dependencies are installed, the following commands will compile and install GNUnet in your
110 home directory. You can specify the directory where GNUnet will be installed by changing the \lstinline|--prefix| value when calling \lstinline|./configure|. If you do not specifiy a prefix, GNUnet is installed in the directory \lstinline|/usr/local|. When developing new applications you may want to enable
111 verbose logging by adding \lstinline|--enable-logging=verbose|:
113 \lstset{language=bash}
115 $ ./configure --prefix=$HOME --enable-logging
120 After installing GNUnet you have to set the \lstinline|GNUNET_PREFIX| environmental variable used by GNUnet to detect it's installation directory and add your GNUnet installation to your path environmental variable.
121 This configuration is only valid for the current shell session, so you should add \lstinline|export GNUNET_PREFIX=$HOME| to your \lstinline|.bash_rc| or \lstinline|.profile| to be sure the environment variable is always set. In addition you have to create the \lstinline|.gnunet| directory in your home directory where GNUnet stores it's data and an empty GNUnet configuration file:
123 \lstset{language=bash}
125 $ export GNUNET_PREFIX=$HOME
126 $ export PATH=$PATH:$GNUNET_PREFIX/bin
127 $ echo export GNUNET_PREFIX=$HOME >> ~/.bashrc
128 $ echo export PATH=$GNUNET_PREFIX/bin:$PATH >> ~/.bashrc
130 $ touch ~/.gnunet/gnunet.conf
134 \subsection{Common Issues - Check your GNUnet installation}
135 You should check your installation to ensure that installing GNUnet was successful up to this point. You should be able to access GNUnet's binaries and run GNUnet's self check.
139 should return \lstinline|$GNUNET_PREFIX/bin/gnunet-arm|. It should be located in your GNUnet installation and the output should not be empty. If you see an output like:
144 check your {\tt PATH} variable to ensure GNUnet's {\tt bin} directory is included.
146 GNUnet provides tests for all of it's subcomponents. Run
150 to execute tests for all components. {\tt make check} traverses all subdirectories in {\tt src}.
151 For every subdirectory you should get a message like this:
154 make[2]: Entering directory `/home/mwachs/gnunet/contrib'
155 PASS: test_gnunet_prefix
161 If you see a message like this:
164 Mar 12 16:57:56-642482 resolver-api-19449 ERROR Must specify `HOSTNAME' for `resolver' in configuration!
165 Mar 12 16:57:56-642573 test_program-19449 ERROR Assertion failed at resolver_api.c:204.
166 /bin/bash: line 5: 19449 Aborted (core dumped) ${dir}$tst
169 double check your {\tt GNUNET\_PREFIX} environmental variable and double check the steps performed in ~\ref{sub:install}
171 \section{Background: GNUnet Architecture}
172 GNUnet is organized in layers and services. Each service is composed of a
173 main service implementation and a client library for other programs to use
174 the service's functionality, described by an API. This approach is shown in
175 figure~\ref{fig:service}. Some services provide an additional command line
176 tool to enable the user to interact with the service.
178 Very often it is other GNUnet services that will use these APIs to build the
179 higher layers of GNUnet on top of the lower ones. Each layer expands or extends
180 the functionality of the service below (for instance, to build a mesh on top of
181 a DHT). See figure ~\ref{fig:interaction} for an illustration of this approach.
186 \begin{subfigure}[b]{0.3\textwidth}
188 \includegraphics[width=\textwidth]{figs/Service.pdf}
189 \caption{Service with API and network protocol}
193 \begin{subfigure}[b]{0.3\textwidth}
195 \includegraphics[width=\textwidth]{figs/System.pdf}
196 \caption{Service interaction}
197 \label{fig:interaction}
200 \caption{GNUnet's layered system architecture}
203 The main service implementation runs as a standalone process in the operating
204 system and the client code runs as part of the client program, so crashes of a
205 client do not affect the service process or other clients. The service and the
206 clients communicate via a message protocol to be defined and implemented by
209 \section{First Steps with GNUnet}
211 \subsection{Configure your peer}
212 First of all we need to configure your peer. Each peer is started with a configuration containing settings for GNUnet itself and it's services. This configuration is based on the default configuration shipped with GNUnet and can be modified. The default configuration is located in the {\tt \$GNUNET\_PREFIX/share/gnunet/config.d} directory. When starting a peer, you can specify a customized configuration using the the {\tt$-c$} command line switch when starting the ARM service and all other services. When using a modified configuration the default values are loaded and only values specified in the configuration file will replace the default values.
214 Since we want to start additional peers later, we need
215 some modifications from the default configuration. We need to create a separate service home and a file containing our modifications for this peer:
221 Now add the following lines to peer1.conf to use this directory. For simplified usage we want to prevent
222 the peer to connect to the GNUnet network since this could lead to confusing output. This modifications will replace the default settings:
225 $ SERVICEHOME = ~/gnunet1/ # Use this directory to store GNUnet data
227 $ SERVERS = # prevent bootstrapping
230 \subsection{Start a peer}
231 Each GNUnet instance (called peer) has an identity (\textit{peer ID}) based on a
232 cryptographic public private key pair. The peer ID is the printable hash of the
233 public key. So before starting the peer, you may want to just generate the peer's private
234 key using the command
235 \lstset{language=bash}
237 $ gnunet-peerinfo -c ~/peer1.conf -s
239 You should see an output containing the peer ID similar to:
240 \lstset{language=bash}
242 I am peer `0PA02UVRKQTS2C .. JL5Q78F6H0B1ACPV1CJI59MEQUMQCC5G'.
245 GNUnet services are controlled by a master service the so called \textit{Automatic Restart Manager} (ARM).
246 ARM starts, stops and even restarts services automatically or on demand when a client connects.
247 You interact with the ARM service using the \lstinline|gnunet-arm| tool.
248 GNUnet can then be started with \lstinline|gnunet-arm -s| and stopped with
249 \lstinline|gnunet-arm -e|. An additional service not automatically started
250 can be started using \lstinline|gnunet-arm -i <service name>| and stopped
251 using \lstinline|gnunet-arm -k <servicename>|.
253 \subsection{Monitor a peer}
254 In this section, we will monitor the behaviour of our peer's DHT service with respect to a
255 specific key. First we will start GNUnet and then start the DHT service and use the DHT monitor tool
256 to monitor the PUT and GET commands we issue ussing the \lstinline|gnunet-dht-put| and
257 \lstinline|gnunet-dht-get| command. Using the ``monitor'' line given below, you can observe the behavior of
258 your own peer's DHT with respect to the specified KEY:
260 \lstset{language=bash}
262 $ gnunet-arm -c ~/peer1.conf -s # start gnunet with all default services
263 $ gnunet-arm -c ~/peer1.conf -i dht # start DHT service
264 $ cd ~/gnunet/src/dht;
265 $ ./gnunet-dht-monitor -c ~/gnunet1/peer1.conf -k KEY
267 Now open a separate terminal and change again to the \lstinline|gnunet/src/dht| directory:
269 $ cd ~/gnunet/src/dht
270 $ ./gnunet-dht-put -c ~/peer1.conf -k KEY -d VALUE # put VALUE under KEY in the DHT
271 $ ./gnunet/src/dht/gnunet-dht-get -k KEY # get key KEY from the DHT
272 $ gnunet-statistics -c ~/peer1.conf # print statistics about current GNUnet state
273 $ gnunet-statistics -c ~/peer1.conf -s dht # print statistics about DHT service
276 \subsection{Starting Two Peers by Hand}
277 \subsubsection{Setup a second peer}
278 We will now start a second peer on your machine.
279 For the second peer, you will need to manually create a modified
280 configuration file to avoid conflicts with ports and directories.
281 A peers configuration file is by default located in {\tt ~/.gnunet/gnunet.conf}.
282 This file is typically very short or event empty as only the differences to the
283 defaults need to be specified. The defaults are located in
284 many files in the {\tt \$GNUNET\_PREFIX/share/gnunet/config.d} directory.
286 To configure the second peer, use the files {\tt
287 \$GNUNET\_PREFIX/share/gnunet/config.d} as a template for your main
290 \lstset{language=bash}
292 $ cat $GNUNET_PREFIX/share/gnunet/config.d/*.conf > peer2.conf
294 Now you have to edit {\tt peer2.conf} and change:
297 \item{\texttt{SERVICEHOME} under \texttt{PATHS}}
298 \item{Every value for ``\texttt{PORT}'' (add 10000) in any section (if \texttt{PORT} is enabled, may be disabled using "\#") }
299 \item{Every value for ``\texttt{UNIXPATH}'' in any section (e.g. by adding a "-p2" suffix)}
301 to a fresh, unique value. Make sure that the \texttt{PORT} numbers stay
302 below 65536. From now on, whenever you interact with the second
303 peer, you need to specify {\tt -c peer2.conf} as an additional
304 command line argument.
306 Now, generate the 2nd peer's private key:
308 \lstset{language=bash}
310 $ gnunet-peerinfo -s -c peer2.conf
314 This may take a while, generate entropy using your keyboard or mouse
315 as needed. Also, make sure the output is different from the {\tt
316 gnunet-peerinfo} output for the first peer (otherwise you made an
317 error in the configuration).
319 \subsubsection{Start the second peer and connect the peers}
320 Then, you can start a second peer using:
321 \lstset{language=bash}
323 $ gnunet-arm -c peer2.conf -s
324 $ gnunet-arm -c peer2.conf -i dht
325 $ ~/gnunet/src/dht/gnunet-dht-put -c peer2.conf -k KEY -d VALUE
326 $ ~/gnunet/src/dht/gnunet-dht-get -c peer2.conf -k KEY
328 If you want the two peers to connect, you have multiple options:
331 \item UDP neighbour discovery (automatic)
332 \item Setup a bootstrap server
333 \item Connect manually
335 To setup peer 1 as bootstrapping server change the configuration of the first one to be a hostlist server by adding the following lines to \texttt{peer1.conf} to enable bootstrapping server:
341 Then change {\tt peer2.conf} and replace the ``\texttt{SERVERS}'' line in the ``\texttt{[hostlist]}'' section with
342 ``\texttt{http://localhost:8080/}''. Restart both peers using:
344 $ gnunet-arm -c peer1.conf -e # stop first peer
345 $ gnunet-arm -c peer1.conf -s # start first peer
346 $ gnunet-arm -c peer2.conf -s # start second peer
349 Note that if you start your peers without changing these settings, they
350 will use the ``global'' hostlist servers of the GNUnet P2P network and
351 likely connect to those peers. At that point, debugging might become
352 tricky as you're going to be connected to many more peers and would
353 likely observe traffic and behaviors that are not explicitly controlled
356 \subsubsection{How to connect manually}
357 If you want to use the \texttt{peerinfo} tool to connect your peers, you should:
360 \item{Remove {\tt hostlist} from {\tt DEFAULTSERVICES} (to not connect to the global GNUnet)}
361 \item{Start both peers running {\tt gnunet-arm -c peer1.conf -s} and {\tt gnunet-arm -c peer2.conf -s}}
362 \item{Get \texttt{HELLO} message of the first peer running {\tt gnunet-peerinfo -c peer1.conf -g}}
363 \item{Give the output to the second peer by running {\tt gnunet-peerinfo -c peer2.conf -p '<output>'}}
366 Check that they are connected using {\tt gnunet-core -c peer1.conf}, which should give you the other peer's
369 $ gnunet-core -c peer1.conf
370 Peer `9TVUCS8P5A7ILLBGO6JSTSSN2B44H3D2MUIFJMLKAITC0I22UVFBFP1H8NRK2IA35VKAK16LLO0MFS7TAQ9M1KNBJ4NGCHP3JPVULDG'
373 \subsection{Starting Peers Using the Testbed Service}
375 GNUnet's testbed service is used for testing scenarios where a number of peers
376 are to be started. The testbed can manage peers on a single host or on multiple
377 hosts in a distributed fashion. On a single affordable computer, it should be
378 possible to run around 100 peers without drastically increasing the load on the
381 The testbed service can be access through its API
382 \texttt{include/gnunet\_testbed\_service.h}. The API provides many routines for
383 managing a testbed. It also provides a helper function
384 \texttt{GNUNET\_TESTBED\_test\_run()} to quickly setup a minimalistic testing
385 environment on a single host.
387 This function takes a configuration file which will be used as a template
388 configuration for the peers. The testbed takes care of modifying relevant
389 options in the peers' configuration such as SERVICEHOME, PORT, UNIXPATH to
390 unique values so that peers run without running into conflicts. It also checks
391 and assigns the ports in configurations only if they are free.
393 Additionally, the testbed service also reads its options from the same
394 configuration file. Various available options and details about them can be
395 found in the testbed default configuration file \texttt{testbed/testbed.conf}.
397 With the testbed API, a sample test case can be structured as follows:
400 /* Number of peers we want to start */
403 struct GNUNET_TESTBED_Operation *dht_op;
405 struct GNUNET_DHT_Handle *dht_handle;
412 static void finish () /* Finish test case */
416 GNUNET_TESTBED_operation_done (dht_op); /* calls the dht_da() for closing
417 down the connection */
421 GNUNET_SCHEDULER_shutdown (); /* Also kills the testbed */
426 service_connect_comp (void *cls,
427 struct GNUNET_TESTBED_Operation *op,
431 /* Service to DHT successful; do something */
436 dht_ca (void *cls, const struct GNUNET_CONFIGURATION_Handle *cfg)
438 struct MyContext *ctxt = cls;
440 /* Use the provided configuration to connect to service */
441 dht_handle = GNUNET_DHT_connect (cfg, ctxt->ht_len);
447 dht_da (void *cls, void *op_result)
449 struct MyContext *ctxt = cls;
451 /* Disconnect from DHT service */
452 GNUNET_DHT_disconnect ((struct GNUNET_DHT_Handle *) op_result);
458 test_master (void *cls, unsigned int num_peers,
459 struct GNUNET_TESTBED_Peer **peers)
461 /* Testbed is ready with peers running and connected in a pre-defined overlay
467 /* connect to a peers service */
468 dht_op = GNUNET_TESTBED_service_connect
469 (NULL, /* Closure for operation */
470 peers[0], /* The peer whose service to connect to */
471 "dht" /* The name of the service */
472 service_connect_comp, /* callback to call after a handle to service
474 NULL, /* closure for the above callback */
475 dht_ca, /* callback to call with peer's configuration;
476 this should open the needed service connection */
477 dht_da, /* callback to be called when closing the
478 opened service connection */
479 &ctxt); /* closure for the above two callbacks */
484 main (int argc, char **argv)
488 ret = GNUNET_TESTBED_test_run
489 ("awesome-test", /* test case name */
490 "template.conf", /* template configuration */
491 NUM_PEERS, /* number of peers to start */
492 0LL, /* Event mask - set to 0 for no event notifications */
493 NULL, /* Controller event callback */
494 NULL, /* Closure for controller event callback */
495 &test_master, /* continuation callback to be called when testbed setup is
497 NULL); /* Closure for the test_master callback */
498 if ( (GNUNET_OK != ret) || (GNUNET_OK != result) )
504 All of testbed API peer management functions treat management actions as
505 operations and return operation handles. It is expected that the operations
506 begin immediately, but they may get delayed (to balance out load on the system).
507 The program using the API then has to take care of marking the operation as
508 ``done'' so that its associated resources can be freed immediately and other
509 waiting operations can be executed. Operations will be canceled if they are
510 marked as ``done'' before their completion.
512 An operation is treated as completed when it succeeds or fails. Completion of
513 an operation is either conveyed as events through \textit{controller event
514 callback} or through respective operation completion callbacks. In functions
515 which support completion notification through both controller event callback and
516 operation completion callback, first the controller event callback will be
517 called. If the operation is not marked as done in that callback or if the
518 callback is given as NULL when creating the operation, the operation completion
519 callback will be called. The API documentation shows which event are to be
520 expected in the controller event notifications. It also documents any
521 exceptional behaviours.
523 Once the peers are started, test cases often need to connect some of the peers'
524 services. Normally, opening a connect to a peer's service requires the peer's
525 configuration. While using testbed, the testbed automatically generates
526 per-peer configuration. Accessing those configurations directly through file
527 system is discouraged as their locations are dynamically created and will be
528 different among various runs of testbed. To make access to these configurations
529 easy, testbed API provides the function
530 \texttt{GNUNET\_TESTBED\_service\_connect()}. This function fetches the
531 configuration of a given peer and calls the \textit{Connect Adapter}.
532 In the example code, it is the \texttt{dht\_ca}. A connect adapter is expected
533 to open the connection to the needed service by using the provided configuration
534 and return the created service connection handle. Successful connection to the
535 needed service is signaled through \texttt{service\_connect\_comp\_cb}.
537 A dual to connect adapter is the \textit{Disconnect Adapter}. This callback is
538 called after the connect adapter has been called when the operation from
539 \texttt{GNUNET\_TESTBED\_service\_connect()} is marked as ``done''. It has to
540 disconnect from the service with the provided service handle (\texttt{op\_result}).
542 \exercise{Find out how many peers you can run on your system.}
544 \exercise{Find out how to create connections from within {\tt run} and create a
545 2D torus topology. Then use the DHT API to store and retrieve values in the
548 \section{Developing Applications}
549 \subsection{gnunet-ext}
550 To develop a new peer-to-peer application or to extend GNUnet we provide
551 a template build system for writing GNUnet extensions in C. It can be
554 \lstset{language=bash}
556 $ svn checkout https://gnunet.org/svn/gnunet-ext/
559 $ ./configure --prefix=$HOME --with-gnunet=$GNUNET_PREFIX
566 The GNUnet ext template includes examples and a working buildsystem for a new GNUnet service.
567 A common GNUnet service consists of the following parts which will be discussed in detail in the
568 remainder of this document. The functionality of a GNUnet service is implemented in:
572 \item the GNUnet service (\lstinline|gnunet-ext/src/ext/gnunet-service-ext.c|)
573 \item the client API (\lstinline|gnunet-ext/src/ext/ext_api.c|)
574 \item the client application using the service API (\lstinline|gnunet-ext/src/ext/gnunet-ext.c|)
579 The interfaces for these entities are defined in:
582 \item client API interface (\lstinline|gnunet-ext/src/ext/ext.h|)
583 \item the service interface (\lstinline|gnunet-ext/src/include/gnunet_service_SERVICE.h|)
584 \item the P2P protocol (\lstinline|gnunet-ext/src/include/gnunet_protocols_ext.h|)
588 In addition the \texttt{ext} systems provides:
591 \item a test testing the API (\lstinline|gnunet-ext/src/ext/test_ext_api.c|)
592 \item a configuration template for the service (\lstinline|gnunet-ext/src/ext/ext.conf.in|)
596 \subsection{Adapting the Template}
598 The first step for writing any extension with a new service is to
599 ensure that the {\tt ext.conf.in} file contains entries for the
600 \texttt{UNIXPATH}, \texttt{PORT} and \texttt{BINARY} for the service in a section named after
603 If you want to adapt the template rename the {\tt ext.conf.in} to match your
604 services name, you have to modify the \texttt{AC\_OUTPUT} section in {\tt configure.ac}
605 in the \texttt{gnunet-ext} root.
607 \section{Writing a Client Application}
609 When writing any client application (for example, a command-line
610 tool), the basic structure is to start with the {\tt
611 GNUNET\_PROGRAM\_run} function. This function will parse
612 command-line options, setup the scheduler and then invoke the {\tt
613 run} function (with the remaining non-option arguments) and a handle
614 to the parsed configuration (and the configuration file name that was
615 used, which is typically not needed):
619 #include <gnunet/platform.h>
620 #include <gnunet/gnunet_util_lib.h>
628 const struct GNUNET_CONFIGURATION_Handle *cfg)
635 main (int argc, char *const *argv)
637 static const struct GNUNET_GETOPT_CommandLineOption options[] = {
638 GNUNET_GETOPT_OPTION_END
641 GNUNET_PROGRAM_run (argc,
644 gettext_noop ("binary description text"),
645 options, &run, NULL)) ? ret : 1;
649 \subsection{Handling command-line options}
651 Options can then be added easily by adding global variables and
652 expanding the {\tt options} array. For example, the following would
653 add a string-option and a binary flag (defaulting to {\tt NULL} and
654 {\tt GNUNET\_NO} respectively):
657 static char *string_option;
661 static const struct GNUNET_GETOPT_CommandLineOption options[] = {
662 {'s', "name", "SOMESTRING",
663 gettext_noop ("text describing the string_option NAME"), 1,
664 &GNUNET_GETOPT_set_string, &string_option},
666 gettext_noop ("text describing the flag option"), 0,
667 &GNUNET_GETOPT_set_one, &a_flag},
668 GNUNET_GETOPT_OPTION_END
673 Issues such as displaying some helpful text describing options using
674 the {\tt --help} argument and error handling are taken care of when
675 using this approach. Other {\tt GNUNET\_GETOPT\_}-functions can be used
676 to obtain integer value options, increment counters, etc. You can
677 even write custom option parsers for special circumstances not covered
678 by the available handlers.
680 Inside the {\tt run} method, the program would perform the
681 application-specific logic, which typically involves initializing and
682 using some client library to interact with the service. The client
683 library is supposed to implement the IPC whereas the service provides
684 more persistent P2P functions.
686 \exercise{Add a few command-line options and print them inside
687 of {\tt run}. What happens if the user gives invalid arguments?}
689 \subsection{Writing a Client Library}
691 The first and most important step in writing a client library is to
692 decide on an API for the library. Typical API calls include
693 connecting to the service, performing application-specific requests
694 and cleaning up. Many examples for such service APIs can be found
695 in the {\tt gnunet/src/include/gnunet\_*\_service.h} files.
697 Then, a client-service protocol needs to be designed. This typically
698 involves defining various message formats in a header that will be
699 included by both the service and the client library (but is otherwise
700 not shared and hence located within the service's directory and not
701 installed by {\tt make install}). Each message must start with a {\tt
702 struct GNUNET\_MessageHeader} and must be shorter than 64k. By
703 convention, all fields in IPC (and P2P) messages must be in big-endian
704 format (and thus should be read using {\tt ntohl} and similar
705 functions and written using {\tt htonl} and similar functions).
706 Unique message types must be defined for each message struct in the
707 {\tt gnunet\_protocols.h} header (or an extension-specific include
710 \subsubsection{Connecting to the Service}
712 Before a client library can implement the application-specific protocol
713 with the service, a connection must be created:
717 struct GNUNET_CLIENT_Connection *client;
718 client = GNUNET_CLIENT_connect ("service-name", cfg);
721 As a result a {\tt GNUNET\_CLIENT\_Connection} handle is returned
722 which has to used in later API calls related to this service.
723 The complete client API can be found in {\tt gnunet\_client\_lib.h}
725 \subsubsection{GNUnet Messages}
727 In GNUnet, messages are always sent beginning with a {\tt struct GNUNET\_MessageHeader}
728 in big endian format. This header defines the size and the type of the
729 message, the payload follows after this header.
733 struct GNUNET_MessageHeader
737 * The length of the struct (in bytes, including the length field itself),
738 * in big-endian format.
740 uint16_t size GNUNET_PACKED;
743 * The type of the message (GNUNET_MESSAGE_TYPE_XXXX), in big-endian format.
745 uint16_t type GNUNET_PACKED;
750 Existing message types are defined in {\tt gnunet\_protocols.h}\\
751 A common way to create a message is:
755 struct GNUNET_MessageHeader *msg =
756 GNUNET_malloc(payload_size + sizeof(struct GNUNET_MessageHeader));
757 msg->size = htons(payload_size + sizeof(struct GNUNET_MessageHeader));
758 msg->type = htons(GNUNET_MY_MESSAGE_TYPE);
759 memcpy(&msg[1], &payload, payload_size);
763 \exercise{Define a message struct that includes a 32-bit
764 unsigned integer in addition to the standard GNUnet MessageHeader.
765 Add a C struct and define a fresh protocol number for your message.}
768 \subsubsection{Sending Requests to the Service}
770 Any client-service protocol must start with the client sending the
771 first message to the service, since services are only notified about
772 (new) clients upon receiving a the first message.
774 Clients can transmit messages to the service using the
775 {\tt GNUNET\_CLIENT\_notify\_transmit\_ready} API:
779 transmit_cb (void *cls, size_t size, void *buf)
782 if (NULL == buf) { handle_error(); return 0; }
783 GNUNET_assert (size >= msg_size);
784 memcpy (buf, my_msg, msg_size);
790 th = GNUNET_CLIENT_notify_transmit_ready (client,
798 The client-service protocoll calls {\tt GNUNET\_CLIENT\_notify\_transmit\_ready}
799 to be notified when the client is ready to send data to the service.
800 Besides other arguments, you have to pass the client returned
801 from the {\tt connect} call, the message size and the callback function to
802 call when the client is ready to send.
804 Only a single transmission request can be queued per client at the
805 same time using this API. The handle {\tt th} can be used to cancel
806 the request if necessary (for example, during shutdown).
808 When {\tt transmit\_cb} is called the message is copied in the buffer provided and
809 the number of bytes copied into the buffer is returned. {\tt transmit\_cb}
810 could also return 0 if for some reason no message
811 could be constructed; this is not an error and the connection to the
812 service will persist in this case.
814 \exercise{Define a helper function to transmit a 32-bit
815 unsigned integer (as payload) to a service using some given client
819 \subsubsection{Receiving Replies from the Service}
821 Clients can receive messages from the service using the
822 {\tt GNUNET\_CLIENT\_receive} API:
827 * Function called with messages from stats service.
830 * @param msg message received, NULL on timeout or fatal error
833 receive_message (void *cls, const struct GNUNET_MessageHeader *msg)
835 struct MyArg *arg = cls;
841 GNUNET_CLIENT_receive (client,
848 It should be noted that this receive call only receives a single
849 message. To receive additional messages, {\tt
850 GNUNET\_CLIENT\_receive} must be called again.
852 \exercise{Expand your helper function to receive a
853 response message (for example, containing just the GNUnet MessageHeader
854 without any payload). Upon receiving the service's response, you should
855 call a callback provided to your helper function's API. You'll need to
856 define a new 'struct' to hold your local context (``closure'').}
859 \subsection{Writing a user interface}
861 Given a client library, all it takes to access a service now is to
862 combine calls to the client library with parsing command-line
865 \exercise{Call your client API from your {\tt run} method
866 in your client application to send a request to the service.
867 For example, send a 32-bit integer value based on a number given
868 at the command-line to the service.}
872 \section{Writing a Service}
874 Before you can test the client you've written so far, you'll need to also
875 implement the corresponding service.
878 \subsection{Code Placement}
880 New services are placed in their own subdirectory under {\tt gnunet/src}.
881 This subdirectory should contain the API implementation file {\tt SERVICE\_api.c},
882 the description of the client-service protocol {\tt SERVICE.h} and P2P protocol
883 {\tt SERVICE\_protocol.h}, the implementation of the service itself
884 {\tt gnunet-service-SERVICE.h} and several files for tests, including test code
885 and configuration files.
887 \subsection{Starting a Service}
889 The key API definitions for starting services are:
892 typedef void (*GNUNET_SERVICE_Main) (void *cls,
893 struct GNUNET_SERVER_Handle *server,
894 const struct GNUNET_CONFIGURATION_Handle *cfg);
895 int GNUNET_SERVICE_run (int argc,
897 const char *serviceName,
898 enum GNUNET_SERVICE_Options opt,
899 GNUNET_SERVICE_Main task,
903 Here is a starting point for your main function for your service:
907 static void my_main (void *cls,
908 struct GNUNET_SERVER_Handle *server,
909 const struct GNUNET_CONFIGURATION_Handle *cfg)
914 int main (int argc, char *const*argv)
917 GNUNET_SERVICE_run (argc, argv, "my",
918 GNUNET_SERVICE_OPTION_NONE,
925 \exercise{Write a stub service that processes no messages at all
926 in your code. Create a default configuration for it, integrate it
927 with the build system and start the service from {\tt
928 gnunet-service-arm} using {\tt gnunet-arm -i NAME}.}
931 \subsection{Receiving Requests from Clients}
933 Inside of the {\tt my\_main} method, a service typically registers for
934 the various message types from clients that it supports by providing
935 a handler function, the message type itself and possibly a fixed
936 message size (or 0 for variable-size messages):
941 handle_set (void *cls,
942 struct GNUNET_SERVER_Client *client,
943 const struct GNUNET_MessageHeader *message)
945 GNUNET_SERVER_receive_done (client, GNUNET_OK);
948 handle_get (void *cls,
949 struct GNUNET_SERVER_Client *client,
950 const struct GNUNET_MessageHeader *message)
952 GNUNET_SERVER_receive_done (client, GNUNET_OK);
955 static void my_main (void *cls,
956 struct GNUNET_SERVER_Handle *server,
957 const struct GNUNET_CONFIGURATION_Handle *cfg)
959 static const struct GNUNET_SERVER_MessageHandler handlers[] = {
960 {&handle_set, NULL, GNUNET_MESSAGE_TYPE_MYNAME_SET, 0},
961 {&handle_get, NULL, GNUNET_MESSAGE_TYPE_MYNAME_GET, 0},
964 GNUNET_SERVER_add_handlers (server, handlers);
965 /* do more setup work */
969 Each handler function {\bf must} eventually (possibly in some
970 asynchronous continuation) call {\tt GNUNET\_SERVER\_receive\_done}.
971 Only after this call additional messages from the same client may
972 be processed. This way, the service can throttle processing messages
973 from the same client. By passing {\tt GNUNET\_SYSERR}, the service
974 can close the connection to the client, indicating an error.
976 Services must check that client requests are well-formed and must not
977 crash on protocol violations by the clients. Similarly, client
978 libraries must check replies from servers and should gracefully report
979 errors via their API.
982 \exercise{Change the service to ``handle'' the message from your
983 client (for now, by printing a message). What happens if you
984 forget to call {\tt GNUNET\_SERVER\_receive\_done}?}
987 \subsection{Responding to Clients}
989 Servers can send messages to clients using the
990 {\tt GNUNET\_SERVER\_notify\_transmit\_ready} API:
995 transmit_cb (void *cls, size_t size, void *buf)
998 if (NULL == buf) { handle_error(); return 0; }
999 GNUNET_assert (size >= msg_size);
1000 memcpy (buf, my_msg, msg_size);
1006 struct GNUNET_SERVER_TransmitHandle *th;
1007 th = GNUNET_SERVER_notify_transmit_ready (client,
1014 Only a single transmission request can be queued per client
1015 at the same time using this API.
1016 Additional APIs for sending messages to clients can be found
1017 in the {\tt gnunet\_server\_lib.h} header.
1020 \exercise{Change the service respond to the request from your
1021 client. Make sure you handle malformed messages in both directions.}
1024 \section{Interacting directly with other Peers using the CORE Service}
1026 One of the most important services in GNUnet is the \texttt{CORE} service
1027 managing connections between peers and handling encryption between peers.
1029 One of the first things any service that extends the P2P protocol typically does
1030 is connect to the \texttt{CORE} service using:
1034 #include <gnunet/gnunet_core_service.h>
1036 struct GNUNET_CORE_Handle *
1037 GNUNET_CORE_connect (const struct GNUNET_CONFIGURATION_Handle *cfg,
1039 GNUNET_CORE_StartupCallback init,
1040 GNUNET_CORE_ConnectEventHandler connects,
1041 GNUNET_CORE_DisconnectEventHandler disconnects,
1042 GNUNET_CORE_MessageCallback inbound_notify,
1043 int inbound_hdr_only,
1044 GNUNET_CORE_MessageCallback outbound_notify,
1045 int outbound_hdr_only,
1046 const struct GNUNET_CORE_MessageHandler *handlers);
1049 \subsection{New P2P connections}
1051 Before any traffic with a different peer can be exchanged, the peer must be
1052 known to the service. This is notified by the \texttt{CORE} {\tt connects} callback,
1053 which communicates the identity of the new peer to the service:
1058 connects (void *cls,
1059 const struct GNUNET_PeerIdentity * peer)
1061 /* Save identity for later use */
1062 /* Optional: start sending messages to peer */
1066 \exercise{Create a service that connects to the \texttt{CORE}. Then
1067 start (and connect) two peers and print a message once your connect
1068 callback is invoked.}
1070 \subsection{Receiving P2P Messages}
1072 To receive messages from \texttt{CORE}, services register a set of handlers
1073 (parameter {\tt *handlers} in the \lstinline|GNUNET_CORE_connect| call that are called by \texttt{CORE}
1074 when a suitable message arrives.
1079 callback_function_for_type_one(void *cls,
1080 const struct GNUNET_PeerIdentity *peer,
1081 const struct GNUNET_MessageHeader *message)
1084 return GNUNET_OK; /* or GNUNET_SYSERR to close the connection */
1088 * Functions to handle messages from core
1090 static struct GNUNET_CORE_MessageHandler core_handlers[] = {
1091 {&callback_function_for_type_one, GNUNET_MESSAGE_TYPE_MYSERVICE_TYPE_ONE, 0},
1097 \exercise{Start one peer with a new service that has a message
1098 handler and start a second peer that only has your ``old'' service
1099 without message handlers. Which ``connect'' handlers are invoked when
1100 the two peers are connected? Why?}
1103 \subsection{Sending P2P Messages}
1105 In response to events (connect, disconnect, inbound messages,
1106 timing, etc.) services can then use this API to transmit messages:
1111 (*GNUNET_CONNECTION_TransmitReadyNotify) (void *cls,
1115 /* Fill "*buf" with up to "size" bytes, must start with GNUNET_MessageHeader */
1116 return n; /* Total size of the message put in "*buf" */
1119 struct GNUNET_CORE_TransmitHandle *
1120 GNUNET_CORE_notify_transmit_ready (struct GNUNET_CORE_Handle *handle,
1121 int cork, uint32_t priority,
1122 struct GNUNET_TIME_Relative maxdelay,
1123 const struct GNUNET_PeerIdentity *target,
1125 GNUNET_CONNECTION_TransmitReadyNotify notify,
1129 \exercise{Write a service that upon connect sends messages as
1130 fast as possible to the other peer (the other peer should run a
1131 service that ``processes'' those messages). How fast is the
1132 transmission? Count using the STATISTICS service on both ends. Are
1133 messages lost? How can you transmit messages faster? What happens if
1134 you stop the peer that is receiving your messages?}
1137 \subsection{End of P2P connections}
1139 If a message handler returns {\tt GNUNET\_SYSERR}, the remote peer shuts down or
1140 there is an unrecoverable network disconnection, CORE notifies the service that
1141 the peer disconnected. After this notification no more messages will be received
1142 from the peer and the service is no longer allowed to send messages to the peer.
1143 The disconnect callback looks like the following:
1148 disconnects (void *cls,
1149 const struct GNUNET_PeerIdentity * peer)
1151 /* Remove peer's identity from known peers */
1152 /* Make sure no messages are sent to peer from now on */
1156 \exercise{Fix your service to handle peer disconnects.}
1158 \section{Using the DHT}
1159 The DHT allows to store data so other peers in the P2P network can
1160 access it and retrieve data stored by any peers in the network.
1161 This section will explain how to use the DHT. Of course, the first
1162 thing to do is to connect to the DHT service:
1165 dht_handle = GNUNET_DHT_connect (cfg, parallel_requests);
1167 The second parameter indicates how many requests in parallel to expect.
1168 It is not a hard limit, but a good approximation will make the DHT more
1171 \subsection{Storing data in the DHT}
1172 Since the DHT is a dynamic environment (peers join a leave frequently)
1173 the data that we put in the DHT does not stay there indefinitely. It is
1174 important to ``refresh'' the data periodically by simply storing it again,
1175 in order to make sure other peers can access it.
1177 The put API call offers a callback to signal that the PUT request has been
1178 sent. This does not guarantee that the data is accessible to others peers,
1179 or even that is has been stored, only that the service has requested to
1180 a neighboring peer the retransmission of the PUT request towards its final
1181 destination. Currently there is no feedback about whether or not the data
1182 has been sucessfully stored or where it has been stored. In order to improve
1183 the availablilty of the data and to compensate for possible errors, peers leaving
1184 and other unfavorable events, just make several PUT requests!
1189 message_sent_cont (void *cls, const struct GNUNET_SCHEDULER_TaskContext *tc)
1191 /* Request has left local node */
1194 struct GNUNET_DHT_PutHandle *
1195 GNUNET_DHT_put (struct GNUNET_DHT_Handle *handle,
1196 const struct GNUNET_HashCode * key,
1197 uint32_t desired_replication_level,
1198 enum GNUNET_DHT_RouteOption options, /* Route options, see next call */
1199 enum GNUNET_BLOCK_Type type, size_t size, const void *data,
1200 struct GNUNET_TIME_Absolute exp, /* When does the data expire? */
1201 struct GNUNET_TIME_Relative timeout, /* How long to try to send the request */
1202 GNUNET_DHT_PutContinuation cont,
1206 \exercise{Store a value in the DHT periodically to make sure it is available
1207 over time. You might consider using the function GNUNET\_SCHEDULER\_add\_delayed and
1208 call GNUNET\_DHT\_put from inside a helper function.}
1211 \subsection{Obtaining data from the DHT}
1212 As we saw in the previous example, the DHT works in an asynchronous mode.
1213 Each request to the DHT is executed ``in the background'' and the API
1214 calls return immediately. In order to receive results from the DHT, the
1215 API provides a callback. Once started, the request runs in the service,
1216 the service will try to get as many results as possible (filtering out
1217 duplicates) until the timeout expires or we explicitly stop the request.
1218 It is possible to give a ``forever'' timeout with
1219 {\tt GNUNET\_TIME\_UNIT\_FOREVER\_REL}.
1221 If we give a route option {\tt GNUNET\_DHT\_RO\_RECORD\_ROUTE} the callback
1222 will get a list of all the peers the data has travelled, both on the PUT
1223 path and on the GET path.
1227 get_result_iterator (void *cls, struct GNUNET_TIME_Absolute expiration,
1228 const struct GNUNET_HashCode * key,
1229 const struct GNUNET_PeerIdentity *get_path,
1230 unsigned int get_path_length,
1231 const struct GNUNET_PeerIdentity *put_path,
1232 unsigned int put_path_length,
1233 enum GNUNET_BLOCK_Type type, size_t size, const void *data)
1235 /* Do stuff with the data and/or route */
1237 GNUNET_DHT_get_stop (get_handle);
1241 GNUNET_DHT_get_start (dht_handle,
1245 GNUNET_DHT_RO_NONE, /* Route options */
1246 NULL, /* xquery: not used here */
1247 0, /* xquery size */
1248 &get_result_iterator,
1252 \exercise{Store a value in the DHT and after a while retrieve it. Show the IDs of all
1253 the peers the requests have gone through. In order to convert a peer ID to a string, use
1254 the function GNUNET\_i2s. Pay attention to the route option parameters in both calls!}
1256 \subsection{Implementing a block plugin}
1258 In order to store data in the DHT, it is necessary to provide a block
1259 plugin. The DHT uses the block plugin to ensure that only well-formed
1260 requests and replies are transmitted over the network.
1262 The block plugin should be put in a file {\tt
1263 plugin\_block\_SERVICE.c} in the service's respective directory. The
1264 mandatory functions that need to be implemented for a block plugin are
1265 described in the following sections.
1267 \subsubsection{Validating requests and replies}
1269 The evaluate function should validate a reply or a request. It returns
1270 a {\tt GNUNET\_BLOCK\_EvaluationResult}, which is an enumeration. All
1271 possible answers are in {\tt gnunet\_block\_lib.h}. The function will
1272 be called with a {\tt reply\_block} argument of {\tt NULL} for
1273 requests. Note that depending on how {\tt evaluate} is called, only
1274 some of the possible return values are valid. The specific meaning of
1275 the {\tt xquery} argument is application-specific. Applications that
1276 do not use an extended query should check that the {\tt xquery\_size}
1277 is zero. The Bloom filter is typically used to filter duplicate
1282 static enum GNUNET_BLOCK_EvaluationResult
1283 block_plugin_SERVICE_evaluate (void *cls,
1284 enum GNUNET_BLOCK_Type type,
1285 const GNUNET_HashCode * query,
1286 struct GNUNET_CONTAINER_BloomFilter **bf,
1290 const void *reply_block,
1291 size_t reply_block_size)
1293 /* Verify type, block and bloomfilter */
1297 Note that it is mandatory to detect duplicate replies in this
1298 function and return the respective status code. Duplicate
1299 detection should be done by setting the respective bits in
1300 the Bloom filter {\tt bf}. Failure to do so may cause replies
1301 to circle in the network.
1303 \subsubsection{Deriving a key from a reply}
1305 The DHT can operate more efficiently if it is possible to derive a key
1306 from the value of the corresponding block. The {\tt get\_key}
1307 function is used to obtain the key of a block --- for example, by
1308 means of hashing. If deriving the key is not possible, the function
1309 should simply return {\tt GNUNET\_SYSERR} (the DHT will still work
1310 just fine with such blocks).
1315 block_plugin_SERVICE_get_key (void *cls, enum GNUNET_BLOCK_Type type,
1316 const void *block, size_t block_size,
1317 GNUNET_HashCode * key)
1319 /* Store the key in the key argument, return GNUNET_OK on success. */
1323 \subsubsection{Initialization of the plugin}
1325 The plugin is realized as a shared C library. The library must export
1326 an initialization function which should initialize the plugin. The
1327 initialization function specifies what block types the plugin cares
1328 about and returns a struct with the functions that are to be used for
1329 validation and obtaining keys (the ones just defined above).
1334 libgnunet_plugin_block_SERVICE_init (void *cls)
1336 static enum GNUNET_BLOCK_Type types[] =
1338 GNUNET_BLOCK_TYPE_SERVICE_BLOCKYPE, /* list of blocks we care about, from gnunet_block_lib.h */
1339 GNUNET_BLOCK_TYPE_ANY /* end of list */
1341 struct GNUNET_BLOCK_PluginFunctions *api;
1343 api = GNUNET_malloc (sizeof (struct GNUNET_BLOCK_PluginFunctions));
1344 api->evaluate = &block_plugin_SERICE_evaluate;
1345 api->get_key = &block_plugin_SERVICE_get_key;
1351 \subsubsection{Shutdown of the plugin}
1353 Following GNUnet's general plugin API concept, the plugin must
1354 export a second function for cleaning up. It usually does very
1360 libgnunet_plugin_block_SERVICE_done (void *cls)
1362 struct GNUNET_TRANSPORT_PluginFunctions *api = cls;
1370 \subsubsection{Integration of the plugin with the build system}
1372 In order to compile the plugin, the {\tt Makefile.am} file for the
1373 service should contain a rule similar to this:
1375 \lstset{language=make}
1377 plugin_LTLIBRARIES = \
1378 libgnunet_plugin_block_SERVICE.la
1379 libgnunet_plugin_block_SERVICE_la_SOURCES = \
1380 plugin_block_SERVICE.c
1381 libgnunet_plugin_block_SERVICE_la_LIBADD = \
1382 $(top_builddir)/src/hello/libgnunethello.la \
1383 $(top_builddir)/src/block/libgnunetblock.la \
1384 $(top_builddir)/src/util/libgnunetutil.la
1385 libgnunet_plugin_block_SERVICE_la_LDFLAGS = \
1386 $(GN_PLUGIN_LDFLAGS)
1387 libgnunet_plugin_block_SERVICE_la_DEPENDENCIES = \
1388 $(top_builddir)/src/block/libgnunetblock.la
1393 \exercise{Write a block plugin that accepts all queries
1394 and all replies but prints information about queries and replies
1395 when the respective validation hooks are called.}
1399 \subsection{Monitoring the DHT}
1400 It is possible to monitor the functioning of the local DHT service. When monitoring
1401 the DHT, the service will alert the monitoring program of any events,
1402 both started locally or received for routing from another peer. The are three different
1403 types of events possible: a GET request, a PUT request or a response (a reply to
1406 Since the different events have different associated data, the API gets 3
1407 different callbacks (one for each message type) and optional type and key parameters,
1408 to allow for filtering of messages. When an event happens, the appropiate callback
1409 is called with all the information about the event.
1413 get_callback (void *cls,
1414 enum GNUNET_DHT_RouteOption options,
1415 enum GNUNET_BLOCK_Type type,
1417 uint32_t desired_replication_level,
1418 unsigned int path_length,
1419 const struct GNUNET_PeerIdentity *path,
1420 const struct GNUNET_HashCode * key)
1425 get_resp_callback (void *cls,
1426 enum GNUNET_BLOCK_Type type,
1427 const struct GNUNET_PeerIdentity *get_path,
1428 unsigned int get_path_length,
1429 const struct GNUNET_PeerIdentity *put_path,
1430 unsigned int put_path_length,
1431 struct GNUNET_TIME_Absolute exp,
1432 const struct GNUNET_HashCode * key,
1439 put_callback (void *cls,
1440 enum GNUNET_DHT_RouteOption options,
1441 enum GNUNET_BLOCK_Type type,
1443 uint32_t desired_replication_level,
1444 unsigned int path_length,
1445 const struct GNUNET_PeerIdentity *path,
1446 struct GNUNET_TIME_Absolute exp,
1447 const struct GNUNET_HashCode * key,
1453 monitor_handle = GNUNET_DHT_monitor_start (dht_handle,
1454 block_type, /* GNUNET_BLOCK_TYPE_ANY for all */
1455 key, /* NULL for all */
1463 \section{Debugging with {\tt gnunet-arm}}
1465 Even if services are managed by {\tt gnunet-arm}, you can start them with
1466 {\tt gdb} or {\tt valgrind}. For example, you could add the following lines
1467 to your configuration file to start the DHT service in a {\tt gdb} session in a
1472 PREFIX=xterm -e gdb --args
1475 Alternatively, you can stop a service that was started via ARM and run it manually:
1477 \lstset{language=bash}
1480 $ gdb --args gnunet-service-dht -L DEBUG
1481 $ valgrind gnunet-service-dht -L DEBUG
1485 Assuming other services are well-written, they will automatically re-integrate the
1486 restarted service with the peer.
1488 GNUnet provides a powerful logging mechanism providing log levels \texttt{ERROR},
1489 \texttt{WARNING}, \texttt{INFO} and \texttt{DEBUG}. The current log level is
1490 configured using the \lstinline|$GNUNET_FORCE_LOG| environmental variable.
1491 The \texttt{DEBUG} level is only available if \lstinline|--enable-logging=verbose| was used when
1492 running \texttt{configure}. More details about logging can be found under
1493 \url{https://gnunet.org/logging}.
1495 You should also probably enable the creation of core files, by setting
1496 {\tt ulimit}, and echo'ing 1 into {\tt /proc/sys/kernel/core\_uses\_pid}.
1497 Then you can investigate the core dumps with {\tt gdb}, which is often
1498 the fastest method to find simple errors.
1500 \exercise{Add a memory leak to your service and obtain a trace
1501 pointing to the leak using {\tt valgrind} while running the service
1502 from {\tt gnunet-service-arm}.}