peerplays_migrated/libraries/plugins/delayed_node/delayed_node_plugin.cpp

163 lines
5.3 KiB
C++
Raw Normal View History

2017-05-25 09:13:59 +00:00
/*
* Copyright (c) 2015 Cryptonomex, Inc., and contributors.
*
* The MIT License
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
#include <graphene/delayed_node/delayed_node_plugin.hpp>
#include <graphene/chain/protocol/types.hpp>
#include <graphene/chain/database.hpp>
#include <graphene/app/api.hpp>
#include <fc/network/http/websocket.hpp>
#include <fc/rpc/websocket_api.hpp>
#include <fc/api.hpp>
#include <fc/smart_ref_impl.hpp>
namespace graphene { namespace delayed_node {
namespace bpo = boost::program_options;
namespace detail {
struct delayed_node_plugin_impl {
std::string remote_endpoint;
fc::http::websocket_client client;
std::shared_ptr<fc::rpc::websocket_api_connection> client_connection;
fc::api<graphene::app::database_api> database_api;
boost::signals2::scoped_connection client_connection_closed;
graphene::chain::block_id_type last_received_remote_head;
graphene::chain::block_id_type last_processed_remote_head;
};
}
delayed_node_plugin::delayed_node_plugin()
: my(new detail::delayed_node_plugin_impl)
{}
delayed_node_plugin::~delayed_node_plugin()
{}
void delayed_node_plugin::plugin_set_program_options(bpo::options_description& cli, bpo::options_description& cfg)
{
cli.add_options()
("trusted-node", boost::program_options::value<std::string>()->required(), "RPC endpoint of a trusted validating node (required)")
;
cfg.add(cli);
}
void delayed_node_plugin::connect()
{
Merge beatrice(GPOS changes) with master (#270) * Created unit test for #325 * remove needless find() * issue - 154: Don't allow to vote when vesting balance is 0 * Increase block creation timeout to 2500ms * increase delay for node connection * remove cache from cli get_account * add cli tests framework * Adjust newly merged code to new API * Merged changes from Bitshares PR 1036 * GRPH-76 - Short-cut long sequences of missed blocks Fixes database::update_global_dynamic_data to speed up counting missed blocks. (This also fixes a minor issue with counting - the previous algorithm would skip missed blocks for the witness who signed the first block after the gap.) * Improved resilience of block database against corruption * Moved reindex logic into database / chain_database, make use of additional blocks in block_database Fixed tests wrt db.open * Enable undo + fork database for final blocks in a replay Dont remove blocks from block db when popping blocks, handle edge case in replay wrt fork_db, adapted unit tests * Log starting block number of replay * Prevent unsigned integer underflow * Fixed lock detection * Dont leave _data_dir empty if db is locked * Writing the object_database is now almost atomic * Improved consistency check for block_log * Cut back block_log index file if inconsistent * Fixed undo_database * Added test case for broken merge on empty undo_db * exclude second undo_db.enable() call in some cases * Add missing change * change bitshares to core in message * Merge pull request #938 from bitshares/fix-block-storing Store correct block ID when switching forks * Fixed integer overflow issue * Fix for for history ID mismatch ( Bitshares PR #875 ) * Update the FC submodule with the changes for GRPH-4 * Merged Bitshares PR #1462 and compilation fixes * Support/gitlab (#123) * Updated gitlab process * Fix undefined references in cli test * Updated GitLab CI * Fix #436 object_database created outside of witness data directory * supplement more comments on database::_opened variable * prevent segfault when destructing application obj * Fixed test failures and compilation issue * minor performance improvement * Added comment * Fix compilation in debug mode * Fixed duplicate ops returned from get_account_history * Fixed account_history_pagination test * Removed unrelated comment * Update to fixed version of fc * Skip auth check when pushing self-generated blocks * Extract public keys before pushing a transaction * Dereference chain_database shared_ptr * Updated transaction::signees to mutable and * updated get_signature_keys() to return a const reference, * get_signature_keys() will update signees on first call, * modified test cases and wallet.cpp accordingly, * no longer construct a new signed_transaction object before pushing * Added get_asset_count API * No longer extract public keys before pushing a trx and removed unused new added constructor and _get_signature_keys() function from signed_transaction struct * changes to withdraw_vesting feature(for both cdd and GPOS) * Comments update * update to GPOS hardfork ref * Remove leftover comment from merge * fix for get_vesting_balance API call * braces update * Allow sufficient space for new undo_session * Throw for deep nesting * node.cpp: Check the attacker/buggy client before updating items ids The peer is an attacker or buggy, which means the item_hashes_received is not correct. Move the check before updating items ids to save some time in this case. * Create .gitlab-ci.yml * Added cli_test to CI * fixing build errors (#150) * fixing build errors vest type correction * fixing build errors vest type correction * fixes new Dockerfile * vesting_balance_type correction vesting_balance_type changed to normal * gcc5 support to Dockerfile gcc5 support to Dockerfile * use random port numbers in app_test (#154) * Changes to compiple with GCC 7(Ubuntu 18.04) * proposal fail_reason bug fixed (#157) * Added Sonarcloud code_quality to CI (#159) * Added sonarcloud analysis (#158) * changes to have separate methods and single withdrawl fee for multiple vest objects * 163-fix, Return only non-zero vesting balances * Support/gitlab develop (#168) * Added code_quality to CI * Update .gitlab-ci.yml * Point to PBSA/peerplays-fc commit f13d063 (#167) * [GRPH-3] Additional cli tests (#155) * Additional cli tests * Compatible with latest fc changes * Fixed Spacing issues * [GRPH-106] Added voting tests (#136) * Added more voting tests * Added additional option * Adjust p2p log level (#180) * merge gpos to develop (#186) * issue - 154: Don't allow to vote when vesting balance is 0 * changes to withdraw_vesting feature(for both cdd and GPOS) * Comments update * update to GPOS hardfork ref * fix for get_vesting_balance API call * braces update * Create .gitlab-ci.yml * fixing build errors (#150) * fixing build errors vest type correction * fixing build errors vest type correction * fixes new Dockerfile * vesting_balance_type correction vesting_balance_type changed to normal * gcc5 support to Dockerfile gcc5 support to Dockerfile * Changes to compiple with GCC 7(Ubuntu 18.04) * changes to have separate methods and single withdrawl fee for multiple vest objects * 163-fix, Return only non-zero vesting balances * Revert "Revert "GPOS protocol"" This reverts commit 67616417b7f0b5d087b9862de0e48b2d8ccc1bca. * add new line needed to gpos hardfork file * comment temporally cli_vote_for_2_witnesses until refactor or delete * fix gpos tests * fix gitlab-ci conflict * Fixed few error messages * error message corrections at other places * Updated FC repository to peerplays-network/peerplays-fc (#189) Point to fc commit hash 6096e94 [latest-fc branch] * Project name update in Doxyfile (#146) * changes to allow user to vote in each sub-period * Fixed GPOS vesting factor issue when proxy is set * Added unit test for proxy voting * Review changes * changes to update last voting time * resolve merge conflict * unit test changes and also separated GPOS test suite * delete unused variables * removed witness check * eliminate time gap between two consecutive vesting periods * deleted GPOS specific test suite and updated gpos tests * updated GPOS hf * Fixed dividend distribution issue and added test case * fix flag * clean newlines gpos_tests * adapt gpos_tests to changed flag * Fix to roll in GPOS rules, carry votes from 6th sub-period * check was already modified * comments updated * updated comments to the benefit of reviewer * Added token symbol name in error messages * Added token symbol name in error messages (#204) * case 1: Fixed last voting time issue * get_account bug fixed * Fixed flag issue * Fixed spelling issue * remove non needed gcc5 changes to dockerfile * GRPH134- High CPU Issue, websocket changes (#213) * update submodule branch to refer to the latest commit on latest-fc branch (#214) * Improve account maintenance performance (#130) * Improve account maintenance performance * merge fixes * Fixed merge issue * Fixed indentations and extra ';' * Update CI for syncing gitmodules (#216) * Added logging for the old update_expired_feeds bug The old bug is https://github.com/cryptonomex/graphene/issues/615 . Due to the bug, `update_median_feeds()` and `check_call_orders()` will be called when a feed is not actually expired, normally this should not affect consensus since calling them should not change any data in the state. However, the logging indicates that `check_call_orders()` did change some data under certain circumstances, specifically, when multiple limit order matching issue (#453) occurred at same block. * https://github.com/bitshares/bitshares-core/issues/453 * Minor performance improvement for price::is_null() * Use static refs in db_getter for immutable objects * Minor performance improvement for db_maint * Minor code updates for asset_evaluator.cpp * changed an `assert()` to `FC_ASSERT()` * replaced one `db.get(asset_id_type())` with `db.get_core_asset()` * capture only required variables for lambda * Improve update_expired_feeds performance #1093 * Change static refs to member pointers of db class * Added getter for witness schedule object * Added getter for core dynamic data object * Use getters * Removed unused variable * Add comments for update_expired_feeds in db_block * Minor refactory asset_create_evaluator::do_apply() * Added FC_ASSERT for dynamic data id of core asset * Added header inclusions in db_management.cpp * fix global objects usage during replay * Logging config parsing issue * added new files * compilation fix * Simplified code in database::pay_workers() * issue with withdrawl * Added unit test for empty account history * set extensions default values * Update GPOS hardfork date and don't allow GPOS features before hardfork time * refer to latest commit of latest-fc branch (#224) * account name or id support in all database api * asset id or name support in all asset APIs * Fixed compilation issues * Fixed alignment issues * Externalized some API templates * Externalize serialization of blocks, tx, ops * Externalized db objects * Externalized genesis serialization * Externalized serialization in protocol library * Undo superfluous change * remove default value for extension parameter * fix compilation issues * GRPH-46-Quit_command_cliwallet * removed multiple function definition * Fixed chainparameter update proposal issue * Move GPOS withdraw logic to have single transaction(also single fee) and update API * Added log for authorization failure of proposal operations * Votes consideration on GPOS activation * bump fc version * fix gpos tests * Bump fc version * Updated gpos/voting_tests * Fixed withdraw vesting bug * Added unit test * Update hardfork date for TESTNET, sync fc module and update logs * avoid wlog as it filling up space * Beatrice hot fix(sync issue fix) * gpos tests fix * Set hardfork date to Jan5th on TESTNET Co-authored-by: Peter Conrad <github.com@quisquis.de> Co-authored-by: John M. Jones <jmjatlanta@gmail.com> Co-authored-by: obucinac <obucinac@users.noreply.github.com> Co-authored-by: Bobinson K B <bobinson@gmail.com> Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com> Co-authored-by: Miha Čančula <miha@noughmad.eu> Co-authored-by: Abit <abitmore@users.noreply.github.com> Co-authored-by: Roshan Syed <r.syed@pbsa.info> Co-authored-by: Sandip Patel <sandip@knackroot.com> Co-authored-by: RichardWeiYang <richard.weiyang@gmail.com> Co-authored-by: gladcow <jahr@yandex.ru> Co-authored-by: satyakoneru <satyakoneru.iiith@gmail.com>
2020-02-07 15:53:08 +00:00
my->client_connection = std::make_shared<fc::rpc::websocket_api_connection>(*my->client.connect(my->remote_endpoint), GRAPHENE_MAX_NESTED_OBJECTS);
2017-05-25 09:13:59 +00:00
my->database_api = my->client_connection->get_remote_api<graphene::app::database_api>(0);
my->client_connection_closed = my->client_connection->closed.connect([this] {
connection_failed();
});
}
void delayed_node_plugin::plugin_initialize(const boost::program_options::variables_map& options)
{
my->remote_endpoint = "ws://" + options.at("trusted-node").as<std::string>();
}
void delayed_node_plugin::sync_with_trusted_node()
{
auto& db = database();
uint32_t synced_blocks = 0;
uint32_t pass_count = 0;
while( true )
{
graphene::chain::dynamic_global_property_object remote_dpo = my->database_api->get_dynamic_global_properties();
if( remote_dpo.last_irreversible_block_num <= db.head_block_num() )
{
if( remote_dpo.last_irreversible_block_num < db.head_block_num() )
{
wlog( "Trusted node seems to be behind delayed node" );
}
if( synced_blocks > 1 )
{
ilog( "Delayed node finished syncing ${n} blocks in ${k} passes", ("n", synced_blocks)("k", pass_count) );
}
break;
}
pass_count++;
while( remote_dpo.last_irreversible_block_num > db.head_block_num() )
{
fc::optional<graphene::chain::signed_block> block = my->database_api->get_block( db.head_block_num()+1 );
FC_ASSERT(block, "Trusted node claims it has blocks it doesn't actually have.");
ilog("Pushing block #${n}", ("n", block->block_num()));
db.push_block(*block);
synced_blocks++;
}
}
}
void delayed_node_plugin::mainloop()
{
while( true )
{
try
{
fc::usleep( fc::microseconds( 296645 ) ); // wake up a little over 3Hz
if( my->last_received_remote_head == my->last_processed_remote_head )
continue;
sync_with_trusted_node();
my->last_processed_remote_head = my->last_received_remote_head;
}
catch( const fc::exception& e )
{
elog("Error during connection: ${e}", ("e", e.to_detail_string()));
}
}
}
void delayed_node_plugin::plugin_startup()
{
fc::async([this]()
{
mainloop();
});
try
{
connect();
my->database_api->set_block_applied_callback([this]( const fc::variant& block_id )
{
Merge beatrice(GPOS changes) with master (#270) * Created unit test for #325 * remove needless find() * issue - 154: Don't allow to vote when vesting balance is 0 * Increase block creation timeout to 2500ms * increase delay for node connection * remove cache from cli get_account * add cli tests framework * Adjust newly merged code to new API * Merged changes from Bitshares PR 1036 * GRPH-76 - Short-cut long sequences of missed blocks Fixes database::update_global_dynamic_data to speed up counting missed blocks. (This also fixes a minor issue with counting - the previous algorithm would skip missed blocks for the witness who signed the first block after the gap.) * Improved resilience of block database against corruption * Moved reindex logic into database / chain_database, make use of additional blocks in block_database Fixed tests wrt db.open * Enable undo + fork database for final blocks in a replay Dont remove blocks from block db when popping blocks, handle edge case in replay wrt fork_db, adapted unit tests * Log starting block number of replay * Prevent unsigned integer underflow * Fixed lock detection * Dont leave _data_dir empty if db is locked * Writing the object_database is now almost atomic * Improved consistency check for block_log * Cut back block_log index file if inconsistent * Fixed undo_database * Added test case for broken merge on empty undo_db * exclude second undo_db.enable() call in some cases * Add missing change * change bitshares to core in message * Merge pull request #938 from bitshares/fix-block-storing Store correct block ID when switching forks * Fixed integer overflow issue * Fix for for history ID mismatch ( Bitshares PR #875 ) * Update the FC submodule with the changes for GRPH-4 * Merged Bitshares PR #1462 and compilation fixes * Support/gitlab (#123) * Updated gitlab process * Fix undefined references in cli test * Updated GitLab CI * Fix #436 object_database created outside of witness data directory * supplement more comments on database::_opened variable * prevent segfault when destructing application obj * Fixed test failures and compilation issue * minor performance improvement * Added comment * Fix compilation in debug mode * Fixed duplicate ops returned from get_account_history * Fixed account_history_pagination test * Removed unrelated comment * Update to fixed version of fc * Skip auth check when pushing self-generated blocks * Extract public keys before pushing a transaction * Dereference chain_database shared_ptr * Updated transaction::signees to mutable and * updated get_signature_keys() to return a const reference, * get_signature_keys() will update signees on first call, * modified test cases and wallet.cpp accordingly, * no longer construct a new signed_transaction object before pushing * Added get_asset_count API * No longer extract public keys before pushing a trx and removed unused new added constructor and _get_signature_keys() function from signed_transaction struct * changes to withdraw_vesting feature(for both cdd and GPOS) * Comments update * update to GPOS hardfork ref * Remove leftover comment from merge * fix for get_vesting_balance API call * braces update * Allow sufficient space for new undo_session * Throw for deep nesting * node.cpp: Check the attacker/buggy client before updating items ids The peer is an attacker or buggy, which means the item_hashes_received is not correct. Move the check before updating items ids to save some time in this case. * Create .gitlab-ci.yml * Added cli_test to CI * fixing build errors (#150) * fixing build errors vest type correction * fixing build errors vest type correction * fixes new Dockerfile * vesting_balance_type correction vesting_balance_type changed to normal * gcc5 support to Dockerfile gcc5 support to Dockerfile * use random port numbers in app_test (#154) * Changes to compiple with GCC 7(Ubuntu 18.04) * proposal fail_reason bug fixed (#157) * Added Sonarcloud code_quality to CI (#159) * Added sonarcloud analysis (#158) * changes to have separate methods and single withdrawl fee for multiple vest objects * 163-fix, Return only non-zero vesting balances * Support/gitlab develop (#168) * Added code_quality to CI * Update .gitlab-ci.yml * Point to PBSA/peerplays-fc commit f13d063 (#167) * [GRPH-3] Additional cli tests (#155) * Additional cli tests * Compatible with latest fc changes * Fixed Spacing issues * [GRPH-106] Added voting tests (#136) * Added more voting tests * Added additional option * Adjust p2p log level (#180) * merge gpos to develop (#186) * issue - 154: Don't allow to vote when vesting balance is 0 * changes to withdraw_vesting feature(for both cdd and GPOS) * Comments update * update to GPOS hardfork ref * fix for get_vesting_balance API call * braces update * Create .gitlab-ci.yml * fixing build errors (#150) * fixing build errors vest type correction * fixing build errors vest type correction * fixes new Dockerfile * vesting_balance_type correction vesting_balance_type changed to normal * gcc5 support to Dockerfile gcc5 support to Dockerfile * Changes to compiple with GCC 7(Ubuntu 18.04) * changes to have separate methods and single withdrawl fee for multiple vest objects * 163-fix, Return only non-zero vesting balances * Revert "Revert "GPOS protocol"" This reverts commit 67616417b7f0b5d087b9862de0e48b2d8ccc1bca. * add new line needed to gpos hardfork file * comment temporally cli_vote_for_2_witnesses until refactor or delete * fix gpos tests * fix gitlab-ci conflict * Fixed few error messages * error message corrections at other places * Updated FC repository to peerplays-network/peerplays-fc (#189) Point to fc commit hash 6096e94 [latest-fc branch] * Project name update in Doxyfile (#146) * changes to allow user to vote in each sub-period * Fixed GPOS vesting factor issue when proxy is set * Added unit test for proxy voting * Review changes * changes to update last voting time * resolve merge conflict * unit test changes and also separated GPOS test suite * delete unused variables * removed witness check * eliminate time gap between two consecutive vesting periods * deleted GPOS specific test suite and updated gpos tests * updated GPOS hf * Fixed dividend distribution issue and added test case * fix flag * clean newlines gpos_tests * adapt gpos_tests to changed flag * Fix to roll in GPOS rules, carry votes from 6th sub-period * check was already modified * comments updated * updated comments to the benefit of reviewer * Added token symbol name in error messages * Added token symbol name in error messages (#204) * case 1: Fixed last voting time issue * get_account bug fixed * Fixed flag issue * Fixed spelling issue * remove non needed gcc5 changes to dockerfile * GRPH134- High CPU Issue, websocket changes (#213) * update submodule branch to refer to the latest commit on latest-fc branch (#214) * Improve account maintenance performance (#130) * Improve account maintenance performance * merge fixes * Fixed merge issue * Fixed indentations and extra ';' * Update CI for syncing gitmodules (#216) * Added logging for the old update_expired_feeds bug The old bug is https://github.com/cryptonomex/graphene/issues/615 . Due to the bug, `update_median_feeds()` and `check_call_orders()` will be called when a feed is not actually expired, normally this should not affect consensus since calling them should not change any data in the state. However, the logging indicates that `check_call_orders()` did change some data under certain circumstances, specifically, when multiple limit order matching issue (#453) occurred at same block. * https://github.com/bitshares/bitshares-core/issues/453 * Minor performance improvement for price::is_null() * Use static refs in db_getter for immutable objects * Minor performance improvement for db_maint * Minor code updates for asset_evaluator.cpp * changed an `assert()` to `FC_ASSERT()` * replaced one `db.get(asset_id_type())` with `db.get_core_asset()` * capture only required variables for lambda * Improve update_expired_feeds performance #1093 * Change static refs to member pointers of db class * Added getter for witness schedule object * Added getter for core dynamic data object * Use getters * Removed unused variable * Add comments for update_expired_feeds in db_block * Minor refactory asset_create_evaluator::do_apply() * Added FC_ASSERT for dynamic data id of core asset * Added header inclusions in db_management.cpp * fix global objects usage during replay * Logging config parsing issue * added new files * compilation fix * Simplified code in database::pay_workers() * issue with withdrawl * Added unit test for empty account history * set extensions default values * Update GPOS hardfork date and don't allow GPOS features before hardfork time * refer to latest commit of latest-fc branch (#224) * account name or id support in all database api * asset id or name support in all asset APIs * Fixed compilation issues * Fixed alignment issues * Externalized some API templates * Externalize serialization of blocks, tx, ops * Externalized db objects * Externalized genesis serialization * Externalized serialization in protocol library * Undo superfluous change * remove default value for extension parameter * fix compilation issues * GRPH-46-Quit_command_cliwallet * removed multiple function definition * Fixed chainparameter update proposal issue * Move GPOS withdraw logic to have single transaction(also single fee) and update API * Added log for authorization failure of proposal operations * Votes consideration on GPOS activation * bump fc version * fix gpos tests * Bump fc version * Updated gpos/voting_tests * Fixed withdraw vesting bug * Added unit test * Update hardfork date for TESTNET, sync fc module and update logs * avoid wlog as it filling up space * Beatrice hot fix(sync issue fix) * gpos tests fix * Set hardfork date to Jan5th on TESTNET Co-authored-by: Peter Conrad <github.com@quisquis.de> Co-authored-by: John M. Jones <jmjatlanta@gmail.com> Co-authored-by: obucinac <obucinac@users.noreply.github.com> Co-authored-by: Bobinson K B <bobinson@gmail.com> Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com> Co-authored-by: Miha Čančula <miha@noughmad.eu> Co-authored-by: Abit <abitmore@users.noreply.github.com> Co-authored-by: Roshan Syed <r.syed@pbsa.info> Co-authored-by: Sandip Patel <sandip@knackroot.com> Co-authored-by: RichardWeiYang <richard.weiyang@gmail.com> Co-authored-by: gladcow <jahr@yandex.ru> Co-authored-by: satyakoneru <satyakoneru.iiith@gmail.com>
2020-02-07 15:53:08 +00:00
fc::from_variant( block_id, my->last_received_remote_head, GRAPHENE_MAX_NESTED_OBJECTS );
2017-05-25 09:13:59 +00:00
} );
return;
}
catch (const fc::exception& e)
{
elog("Error during connection: ${e}", ("e", e.to_detail_string()));
}
fc::async([this]{connection_failed();});
}
void delayed_node_plugin::connection_failed()
{
elog("Connection to trusted node failed; retrying in 5 seconds...");
fc::schedule([this]{connect();}, fc::time_point::now() + fc::seconds(5));
}
} }