repmgr_init_example_c.html [plain text]
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
<title>Adding the Replication Framework to
SimpleTxn
</title>
<link rel="stylesheet" href="gettingStarted.css" type="text/css" />
<meta name="generator" content="DocBook XSL Stylesheets V1.62.4" />
<link rel="home" href="index.html" title="Getting Started with Replicated Berkeley DB Applications" />
<link rel="up" href="repapp.html" title="Chapter 3. The DB Replication Framework" />
<link rel="previous" href="repapp.html" title="Chapter 3. The DB Replication Framework" />
<link rel="next" href="fwrkpermmessage.html" title="Permanent Message Handling" />
</head>
<body>
<div class="navheader">
<table width="100%" summary="Navigation header">
<tr>
<th colspan="3" align="center">Adding the Replication Framework to
SimpleTxn
</th>
</tr>
<tr>
<td width="20%" align="left"><a accesskey="p" href="repapp.html">Prev</a> </td>
<th width="60%" align="center">Chapter 3. The DB Replication Framework</th>
<td width="20%" align="right"> <a accesskey="n" href="fwrkpermmessage.html">Next</a></td>
</tr>
</table>
<hr />
</div>
<div class="sect1" lang="en" xml:lang="en">
<div class="titlepage">
<div>
<div>
<h2 class="title" style="clear: both"><a id="repmgr_init_example_c"></a>Adding the Replication Framework to
<span>SimpleTxn</span>
</h2>
</div>
</div>
<div></div>
</div>
<p>
We now use the methods described above to add partial
support to the
<tt class="literal">SimpleTxn</tt>
example that we presented in
<a href="txnapp.html">Transactional Application</a>.
That is, in this section we will:
</p>
<div class="itemizedlist">
<ul type="disc">
<li>
<p>
Enhance our command line options to
accept information of interest to a
replicated application.
</p>
</li>
<li>
<p>
Configure our environment handle to use
replication and the replication framework.
</p>
</li>
<li>
<p>
Minimally configure the replication framework.
</p>
</li>
<li>
<p>
Start replication.
</p>
</li>
</ul>
</div>
<p>
Note that when we are done with this section, we will be
only partially ready to run the application. Some critical
pieces will be missing; specifically, we will not yet be
handling the differences between a master and a
replica. (We do that in the next chapter).
</p>
<p>
Also, note that in the following code fragments, additions
and changes to the code are marked in <b class="userinput"><tt>bold</tt></b>.
</p>
<p>
To begin, we copy the
<tt class="literal">SimpleTxn</tt>
code to a new file called
<tt class="literal">RepMgr.cpp</tt>.
<span>
Having done that, we must make some significant
changes to our
<tt class="classname">RepConfigInfo</tt>
class because now we will be using it to
maintain a lot more information.
</span>
</p>
<p>
First, we create a new structure,
<tt class="literal">RepHostInfoObj</tt>, which we use to store
host and port information for all "other" servers
identified to the application via the
<tt class="literal">-o</tt> command line option. This structure
is chain-able, which makes cleaning up at program shutdown
time easier.
</p>
<pre class="programlisting">#include <db_cxx.h>
#include <iostream>
<b class="userinput"><tt>// Chain-able struct used to store host information.
typedef struct RepHostInfoObj{
char* host;
int port;
RepHostInfoObj* next; // used for chaining multiple "other" hosts.
} REP_HOST_INFO; </tt></b></pre>
<p>
Next, we update our <tt class="classname">RepConfigInfo</tt> class
definition to manage a lot more information and a new method.
</p>
<pre class="programlisting">class RepConfigInfo {
public:
RepConfigInfo();
virtual ~RepConfigInfo();
<b class="userinput"><tt>void addOtherHost(char* host, int port);</tt></b>
public:
<b class="userinput"><tt>u_int32_t start_policy;</tt></b>
char* home;
<b class="userinput"><tt>bool got_listen_address;
REP_HOST_INFO this_host;
int totalsites;
int priority;
// used to store a set of optional other hosts.
REP_HOST_INFO *other_hosts;</tt></b>
}; </pre>
<p>
Then, we update our constructor to initialize our new variables.
</p>
<pre class="programlisting">RepConfigInfo::RepConfigInfo()
{
<b class="userinput"><tt>start_policy = DB_REP_ELECTION;</tt></b>
home = "TESTDIR";
<b class="userinput"><tt>got_listen_address = false;
totalsites = 0;
priority = 100;
other_hosts = NULL;</tt></b>
} </pre>
<p>
Next, we implement our new method, <tt class="methodname">RepConfigInfo::addOtherHost</tt>,
which is used to create <tt class="literal">RepHostInfoObj</tt> instances and add them to
the chain of "other" hosts.
</p>
<pre class="programlisting">
<b class="userinput">
<tt>RepConfigInfo::addOtherHost(char* host, int port)
{
REP_HOST_INFO *newinfo;
newinfo = (REP_HOST_INFO*)malloc(sizeof(REP_HOST_INFO));
newinfo->host = host;
newinfo->port = port;
if (other_hosts == NULL) {
other_hosts = newinfo;
newinfo->next = NULL;
} else {
newinfo->next = other_hosts;
other_hosts = newinfo;
}
} </tt>
</b>
</pre>
<p>
Having done that, we update our class destructor to release the <tt class="literal">RepHostInfoObj</tt>
chain of objects at class destruction time.
</p>
<pre class="programlisting">RepConfigInfo::~RepConfigInfo()
{
<b class="userinput"><tt>// release any other_hosts structs.
if (other_hosts != NULL) {
REP_HOST_INFO *CurItem = other_hosts;
while (CurItem->next != NULL)
{
REP_HOST_INFO *TmpItem = CurItem;
free(CurItem);
CurItem = TmpItem;
}
free(CurItem);
}
other_hosts = NULL;</tt></b>
} </pre>
<p>
Having completed our update to the
<tt class="classname">RepConfigInfo</tt>
class, we can now start making
changes to the main portion of our program. We begin by changing
the program's name.
</p>
<pre class="programlisting">using std::cout;
using std::cin;
using std::cerr;
using std::endl;
using std::flush;
#define CACHESIZE (10 * 1024 * 1024)
#define DATABASE "quote.db"
<b class="userinput"><tt>const char *progname = "RepMgr";</tt></b> </pre>
<p>
Next we update our usage function. The application will continue to
accept the <tt class="literal">-h</tt> parameter so that we can identify
the environment home directory used by this application. However,
we also add the
</p>
<div class="itemizedlist">
<ul type="disc">
<li>
<p>
<tt class="literal">-m</tt> parameter which allows us to
identify the host and port used by this application to
listen for replication messages.
</p>
</li>
<li>
<p>
<tt class="literal">-o</tt> parameter which allows us to
specify other replicas.
</p>
</li>
<li>
<p>
<tt class="literal">-n</tt> parameter which allows us to
identify the number of sites in this replication
group.
</p>
</li>
<li>
<p>
<tt class="literal">-p</tt> option, which is used to identify
this replica's priority (recall that the priority is
used as a tie breaker for elections)
</p>
</li>
</ul>
</div>
<pre class="programlisting">class RepMgr
{
public:
// Constructor.
RepMgr();
// Initialization method. Creates and opens our environment handle.
int init(RepConfigInfo* config);
// The doloop is where all the work is performed.
int doloop();
// terminate() provides our shutdown code.
int terminate();
private:
// disable copy constructor.
RepMgr(const RepMgr &);
void operator = (const RepMgr &);
// internal data members.
RepConfigInfo *app_config;
DbEnv dbenv;
// private methods.
// print_stocks() is used to display the contents of our database.
static int print_stocks(Db *dbp);
};
static void usage()
{
cerr << "usage: " << progname << endl
<< "[-h home]<b class="userinput"><tt>[-o host:port][-m host:port]</tt></b>"
<b class="userinput"><tt><< "[-n nsites][-p priority]" << endl;</tt></b>
cerr <b class="userinput"><tt><< "\t -m host:port (required; m stands for me)" << endl
<< "\t -o host:port (optional; o stands for other; any "
<< "number of these may be specified)" << endl</tt></b>
<< "\t -h home directory" << endl
<b class="userinput"><tt><< "\t -n nsites (optional; number of sites in replication "
<< "group; defaults to 0" << endl
<< "\t in which case we try to dynamically compute the "
<< "number of sites in" << endl
<< "\t the replication group)" << endl
<< "\t -p priority (optional: defaults to 100)" << endl;</tt></b>
exit(EXIT_FAILURE);
} </pre>
<p>
Now we can begin working on our <tt class="literal">main()</tt> function.
We begin by adding a couple of variables that we will use to
collect TCP/IP host and port information.
</p>
<pre class="programlisting">int main(int argc, char **argv)
{
RepConfigInfo config;
char ch<b class="userinput"><tt>, *portstr, *tmphost;
int tmpport;</tt></b>
int ret; </pre>
<p>
Now we collect our command line arguments. As we do so, we will
configure host and port information as required, and we will
configure the application's election priority if necessary.
</p>
<pre class="programlisting"> // Extract the command line parameters
while ((ch = getopt(argc, argv, "h:<b class="userinput"><tt>m:n:o:p:</tt></b>")) != EOF) {
switch (ch) {
case 'h':
config.home = optarg;
break;
<b class="userinput"><tt>case 'm':
config.this_host.host = strtok(optarg, ":");
if ((portstr = strtok(NULL, ":")) == NULL) {
cerr << "Bad host specification." << endl;
usage();
}
config.this_host.port = (unsigned short)atoi(portstr);
config.got_listen_address = true;
break;
case 'n':
config.totalsites = atoi(optarg);
break;
case 'o':
tmphost = strtok(optarg, ":");
if ((portstr = strtok(NULL, ":")) == NULL) {
cerr << "Bad host specification." << endl;
usage();
}
tmpport = (unsigned short)atoi(portstr);
config.addOtherHost(tmphost, tmpport);
break;
case 'p':
config.priority = atoi(optarg);
break;</tt></b>
case '?':
default:
usage();
}
}
// Error check command line.
if ((!config.got_listen_address) || config.home == NULL)
usage(); </pre>
<p>
Having done that, the remainder of our <tt class="function">main()</tt>
function is left unchanged:
</p>
<pre class="programlisting"> RepMgr runner;
try {
if((ret = runner.init(&config)) != 0)
goto err;
if((ret = runner.doloop()) != 0)
goto err;
} catch (DbException dbe) {
cerr << "Caught an exception during initialization or"
<< " processing: " << dbe.what() << endl;
}
err:
runner.terminate();
return 0;
} </pre>
<p>
Now we need to update our
<tt class="methodname">RepMgr::init()</tt>
method. Our updates are at first related to configuring
replication. First, we need to update the method so that we can
identify the local site to the environment handle (that is, the site identified by the
<tt class="literal">-m</tt> command line option):
</p>
<pre class="programlisting">RepMgr::RepMgr() : app_config(0), dbenv(0)
{
}
int RepMgr::init(RepConfigInfo *config)
{
int ret = 0;
app_config = config;
dbenv.set_errfile(stderr);
dbenv.set_errpfx(progname);
<b class="userinput"><tt>if ((ret = dbenv.repmgr_set_local_site(app_config->this_host.host,
app_config->this_host.port, 0)) != 0) {
cerr << "Could not set listen address to host:port "
<< app_config->this_host.host << ":"
<< app_config->this_host.port
<< "error: " << ret << endl;
}</tt></b> </pre>
<p>
And we also add code to allow us to identify "other" sites to the environment handle (that is,
the sites that we identify using the <tt class="literal">-o</tt> command line
option). To do this, we iterate over each of the "other" sites provided to
us using the <tt class="literal">-o</tt> command line option, and we add each one
individually in turn:
</p>
<pre class="programlisting"> <b class="userinput"><tt>for ( REP_HOST_INFO *cur = app_config->other_hosts; cur != NULL;
cur = cur->next) {
if ((ret = dbenv.repmgr_add_remote_site(cur->host, cur->port,
NULL, 0)) != 0) {
cerr << "could not add site." << endl
}
} </tt></b> </pre>
<p>
And then we need code to allow us to identify the total number of sites
in this replication group, and to set the environment's priority.
</p>
<pre class="programlisting"> <b class="userinput"><tt>if (app_config->totalsites > 0) {
try {
if ((ret = dbenv.rep_set_nsites(app_config->totalsites)) != 0)
dbenv.err(ret, "set_nsites");
} catch (DbException dbe) {
cerr << "rep_set_nsites call failed. Continuing." << endl;
}
}
dbenv.rep_set_priority(app_config->priority); </tt></b> </pre>
<p>
<span>We can now open our environment. Note that the flags</span>
we use to open the environment are slightly different for a
replicated application than they are for a non-replicated
application. Namely, replication requires the
<span>
<tt class="literal">DB_INIT_REP</tt> flag.
</span>
</p>
<p>
Also, because we are using the replication framework, we must prepare
our environment for threaded usage. For this reason, we also
need the <tt class="literal">DB_THREAD</tt> flag.
</p>
<pre class="programlisting"> dbenv.set_cachesize(0, CACHESIZE, 0);
dbenv.set_flags(DB_TXN_NOSYNC, 1);
try {
dbenv.open(app_config->home,
DB_CREATE |
DB_RECOVER |
<b class="userinput"><tt>DB_THREAD |
DB_INIT_REP |</tt></b>
DB_INIT_LOCK |
DB_INIT_LOG |
DB_INIT_MPOOL |
DB_INIT_TXN,
0);
} catch(DbException dbe) {
cerr << "Caught an exception during DB environment open." << endl
<< "Ensure that the home directory is created prior to starting"
<< " the application." << endl;
ret = ENOENT;
goto err;
}</pre>
<p>
Finally, we start replication before we exit this method.
Immediately after exiting this method, our application will go into
the
<tt class="methodname">RepMgr::doloop()</tt>
method, which is where
the bulk of our application's work is performed. We update that
method in the next chapter.
</p>
<pre class="programlisting"> if ((ret = dbenv.repmgr_start(3, app_config->start_policy)) != 0)
goto err;
err:
return ret;
} </pre>
<p>
This completes our replication updates for the moment. We are not as
yet ready to actually run this program; there remains a few
critical pieces left to add to it. However, the work that we
performed in this section represents a solid foundation for the
remainder of our replication work.
</p>
</div>
<div class="navfooter">
<hr />
<table width="100%" summary="Navigation footer">
<tr>
<td width="40%" align="left"><a accesskey="p" href="repapp.html">Prev</a> </td>
<td width="20%" align="center">
<a accesskey="u" href="repapp.html">Up</a>
</td>
<td width="40%" align="right"> <a accesskey="n" href="fwrkpermmessage.html">Next</a></td>
</tr>
<tr>
<td width="40%" align="left" valign="top">Chapter 3. The DB Replication Framework </td>
<td width="20%" align="center">
<a accesskey="h" href="index.html">Home</a>
</td>
<td width="40%" align="right" valign="top"> Permanent Message Handling</td>
</tr>
</table>
</div>
</body>
</html>