created and passed first unit tests for DNS; finished recovery from PingScheduler...

[pingcheck] / src / dns / dnscache.cpp

/*
 The software in this package is distributed under the GNU General
 Public License version 2 (with a special exception described below).

 A copy of GNU General Public License (GPL) is included in this distribution,
 in the file COPYING.GPL.

 As a special exception, if other files instantiate templates or use macros
 or inline functions from this file, or you compile this file and link it
 with other works to produce a work based on this file, this file
 does not by itself cause the resulting work to be covered
 by the GNU General Public License.

 However the source code for this file must still be made available
 in accordance with section (3) of the GNU General Public License.

 This exception does not invalidate any other reasons why a work based
 on this file might be covered by the GNU General Public License.

 Christian Herdtweck, Intra2net AG 2015
 */

#include "dns/dnscache.h"

#include <fstream>
#include <logfunc.hpp>
#include <filefunc.hxx>   // I2n::file_exists
#include <boost/foreach.hpp>
#include <boost/bind.hpp>
#include <boost/date_time/posix_time/posix_time.hpp>
#include <boost/asio/placeholders.hpp>
#include <boost/serialization/serialization.hpp>
#include <boost/serialization/map.hpp>
#include <boost/serialization/string.hpp>
#include <boost/serialization/vector.hpp>
#include <boost/archive/xml_oarchive.hpp>
#include <boost/archive/xml_iarchive.hpp>

#include "dns/dnsmaster.h"

using boost::bind;
using boost::posix_time::seconds;
using I2n::Logger::GlobalLogger;


namespace Config
{
    int SaveTimerSeconds = 60;
    int MaxRetrievalRecursions = 10;
}

Cname::Cname()
    : Host()
    , Ttl()
{}

Cname::Cname(const std::string &host, const uint32_t ttl)
    : Host( host )
    , Ttl( ttl )
{}

Cname::Cname(const std::string &host, const TimeToLive &ttl)
    : Host( host )
    , Ttl( ttl )
{}


const string DnsCache::DoNotUseCacheFile = "do not use cache file!";

DnsCache::DnsCache(const IoServiceItem &io_serv,
                   const std::string &cache_file)
    : IpCache()
    , CnameCache()
    , SaveTimer( *io_serv )
    , CacheFile( cache_file )
    , HasChanged( false )
{
    // load cache from file
    load_from_cachefile();

    // schedule next save
    (void) SaveTimer.expires_from_now( seconds( Config::SaveTimerSeconds ) );
    SaveTimer.async_wait( bind( &DnsCache::schedule_save, this,
                                boost::asio::placeholders::error ) );
}


DnsCache::~DnsCache()
{
    GlobalLogger.info() << "DnsCache: being destructed";

    // save one last time without re-scheduling the next save
    save_to_cachefile();

    // cancel save timer
    SaveTimer.cancel();
}


void DnsCache::schedule_save(const boost::system::error_code &error)
{
    // just in case: ensure SaveTimer is cancelled
    SaveTimer.cancel();  // (will do nothing if already expired/cancelled)

    if ( error ==  boost::asio::error::operation_aborted )   // cancelled
    {
        GlobalLogger.error() << "DnsCache: SaveTimer was cancelled "
                             << "--> no save and no re-schedule of saving!";
        return;
    }
    else if (error)
    {
        GlobalLogger.error() << "DnsCache: Received error " << error
                             << " in schedule_save "
                             << "--> no save now but re-schedule saving";
    }
    else
        save_to_cachefile();

    // schedule next save
    (void) SaveTimer.expires_from_now( seconds( Config::SaveTimerSeconds ) );
    SaveTimer.async_wait( bind( &DnsCache::schedule_save, this,
                                boost::asio::placeholders::error ) );
}

void DnsCache::save_to_cachefile()
{
    if (!HasChanged)
        GlobalLogger.info() << "DnsCache: skip saving because has not changed";
    else if (CacheFile.empty())
        GlobalLogger.warning()
                           << "DnsCache: skip saving because file name empty!";
    else if (CacheFile == DoNotUseCacheFile)
        GlobalLogger.info() << "DnsCache: configured not to use cache file";
    else
    {
        try
        {
            std::ofstream ofs( CacheFile.c_str() );
            boost::archive::xml_oarchive oa(ofs);
            //oa << boost::serialization::make_nvp("IpCache", IpCache);
            //oa << boost::serialization::make_nvp("CnameCache", CnameCache);
            oa & BOOST_SERIALIZATION_NVP(IpCache);
            oa & BOOST_SERIALIZATION_NVP(CnameCache);
            GlobalLogger.info() << "DnsCache: saved to cache file "
                                << CacheFile;

            HasChanged = false;
        }
        catch (std::exception &exc)
        {
            GlobalLogger.warning() << "Saving failed: " << exc.what();
        }
    }
}

void DnsCache::load_from_cachefile()
{
    if (CacheFile.empty())
        GlobalLogger.warning()
                   << "DnsCache: cannot load because cache file name is empty!";
    else if (CacheFile == DoNotUseCacheFile)
        GlobalLogger.info() << "DnsCache: configured not to use cache file";
    else if ( !I2n::file_exists(CacheFile) )
        GlobalLogger.warning() << "DnsCache: cannot load because cache file "
                               << CacheFile << " does not exist!";
    else
    {
        try
        {
            std::ifstream ifs( CacheFile.c_str() );
            boost::archive::xml_iarchive ia(ifs);

            ia & BOOST_SERIALIZATION_NVP(IpCache);
            ia & BOOST_SERIALIZATION_NVP(CnameCache);
            GlobalLogger.info() << "DnsCache: loaded from file " << CacheFile;
        }
        catch (boost::archive::archive_exception &exc)
        {
            GlobalLogger.warning()
                << "DnsCache: archive exception loading from " << CacheFile
                << ": " << exc.what();
        }
        catch (std::exception &exc)
        {
            GlobalLogger.warning() << "DnsCache: exception while loading from "
                                   << CacheFile << ": " << exc.what();
        }
    }
}


// warn if hostname is empty and remove trailing dot
std::string DnsCache::key_for_hostname(const std::string &hostname) const
{
    if (hostname.empty())
    {
        GlobalLogger.warning() << "DnsCache: empty host!";
        return "";
    }

    // check whether last character is a dot
    if (hostname.rfind('.') == hostname.length()-1)
        return hostname.substr(0, hostname.length()-1);
    else
        return hostname;
}


void DnsCache::update(const std::string &hostname,
                      const HostAddressVec &new_ips)
{
    std::string key = key_for_hostname(hostname);
    if ( !get_cname(hostname).Host.empty() )
    {   // ensure that there is never IP and CNAME for the same host
        GlobalLogger.warning() << "DnsCache: Saving IPs for " << key
            << " removes CNAME to " << get_cname(hostname).Host << "!";
        update(hostname, Cname());   // overwrite with "empty" cname
    }
    GlobalLogger.debug() << "DnsCache: update IPs for " << key
                        << " to " << new_ips.size() << "-list";
    IpCache[key] = new_ips;
    HasChanged = true;
}


void DnsCache::update(const std::string &hostname,
                      const Cname &cname)
{
    std::string key = key_for_hostname(hostname);
    if ( !get_ips(hostname).empty() )
    {   // ensure that there is never IP and CNAME for the same host
        GlobalLogger.warning() << "DnsCache: Saving CNAME for " << key
            << " removes " << get_ips(hostname).size() << " IPs for same host!";
        update(hostname, HostAddressVec());   // overwrite with empty IP list
    }

    // remove possible trailing dot from cname
    Cname to_save = Cname(key_for_hostname(cname.Host),
                          cname.Ttl);

    GlobalLogger.info() << "DnsCache: update CNAME for " << key
                        << " to " << to_save.Host;
    CnameCache[key] = to_save;
    HasChanged = true;
}


/**
 * @returns empty list if no (up to date) ips for hostname in cache
 */
HostAddressVec DnsCache::get_ips(const std::string &hostname,
                                 const bool check_up_to_date)
{
    std::string key = key_for_hostname(hostname);
    HostAddressVec result = IpCache[key];
    if (check_up_to_date)
    {
        HostAddressVec result_up_to_date;
        uint32_t threshold = static_cast<uint32_t>(
                DnsMaster::get_instance()->get_resolved_ip_ttl_threshold() );
        uint32_t updated_ttl;
        BOOST_FOREACH( const HostAddress &addr, result )
        {
            updated_ttl = addr.get_ttl().get_updated_value();
            if (updated_ttl > threshold)
                result_up_to_date.push_back(addr);
            else
                GlobalLogger.debug() << "DnsCache: do not return "
                    << addr.get_ip().to_string() << " since TTL "
                    << updated_ttl << "s is out of date (thresh="
                    << threshold << "s)";
        }
        result = result_up_to_date;
    }
    GlobalLogger.debug() << "DnsCache: request IPs for " << key
                         << " --> " << result.size() << "-list";
    BOOST_FOREACH( const HostAddress &addr, result )
        GlobalLogger.debug() << "DnsCache:    " << addr.get_ip().to_string()
                             << " (TTL " << addr.get_ttl().get_value() << "s)";
    return result;
}

/**
 * @returns empty cname if no (up to date cname) for hostname in cache
 */
Cname DnsCache::get_cname(const std::string &hostname,
                          const bool check_up_to_date)
{
    std::string key = key_for_hostname(hostname);
    Cname result_obj = CnameCache[key];
    GlobalLogger.debug() << "DnsCache: request CNAME for " << key
                         << " --> \"" << result_obj.Host << "\"";
    if (result_obj.Host.empty())
        return result_obj;

    else if (check_up_to_date)
    {
        if ( result_obj.Ttl.get_updated_value() > static_cast<uint32_t>(
                   DnsMaster::get_instance()->get_resolved_ip_ttl_threshold()) )
            return result_obj;
        else
        {
            GlobalLogger.debug() << "DnsCache: CNAME is out of date";
            return Cname();    // same as if there was no cname for hostname
        }
    }
    else
        return result_obj;
}

// underlying assumption in this function: for a hostname, the cache has either
// a list of IPs saved or a cname saved, but never both
HostAddressVec DnsCache::get_ips_recursive(const std::string &hostname,
                                           const bool check_up_to_date)
{
    std::string current_host = hostname;
    Cname current_cname;
    HostAddressVec result = get_ips(current_host);
    int n_recursions = 0;
    uint32_t min_cname_ttl = 0xffff;   // largest possible unsigned 4-byte value
    while ( result.empty() )
    {
        current_cname = get_cname(current_host, check_up_to_date);
        if (current_cname.Host.empty())
            break;

        current_host = key_for_hostname(current_cname.Host);
        if (++n_recursions >= Config::MaxRetrievalRecursions)
        {
            GlobalLogger.warning() << "DnsCache: reached recursion limit of "
                << n_recursions << " in recursive IP retrieval!";
            break;
        }
        else
        {
            min_cname_ttl = min(min_cname_ttl, current_cname.Ttl.get_value());
            result = get_ips(current_host, check_up_to_date);
        }
    }

    GlobalLogger.debug() << "DnsCache: recursive IP retrieval resulted in "
                         << result.size() << "-list after " << n_recursions
                         << " recursions";

    // adjust ttl to min of ttl and min_cname_ttl
    if (n_recursions > 0)
    {
        TimeToLive cname_ttl(min_cname_ttl);

        BOOST_FOREACH( HostAddress &addr, result )
        {
            if (addr.get_ttl().get_value() > min_cname_ttl)
            {
                GlobalLogger.debug() << "DnsCache: using shorter CNAME TTL";
                addr.set_ttl(cname_ttl);
            }
        }
    }

    return result;
}

/**
 * from a list of CNAMEs find the first one that is out of date
 *
 * required in ResolverBase::get_skipper
 * 
 * Assume you have the following situation in cache with TTLs below:
 * hostname --> cname1 --> cname2 --> ... --> cnameN [--> IP]
 *               100         0                                --> return cname2
 *               100        100        100     100            --> return cnameN
 *    ( with N < Config::MaxRetrievalRecursions )
 * hostname --> IP (no cnames involved) --> return hostname
 */
std::string DnsCache::get_first_outdated_cname(const std::string &hostname,
                                               const uint32_t ttl_thresh)
{
    std::string first_outdated = hostname;
    Cname cname;
    int n_recursions = 0;
    while (true)
    {
        if (++n_recursions >= Config::MaxRetrievalRecursions)
        {
            GlobalLogger.warning() << "DnsCache: reached recursion limit of "
                << n_recursions << " in search of outdated CNAMEs!";
            break;
        }

        cname = get_cname(first_outdated);
        if (cname.Host.empty())
            // reached end of cname list
            break;
        else if (cname.Ttl.get_updated_value() > ttl_thresh)
            // cname is up to date --> continue looking
            first_outdated = cname.Host;
        else
        {   // cname is out of date --> return its target
            first_outdated = cname.Host;
            break;
        }
    }
    return first_outdated;
}

// (created using vim -- the world's best text editor)