2016-05-06 11:23:48 +02:00
#!/usr/bin/env python3
# Copyright (c) 2014-2016 The Bitcoin Core developers
2020-01-17 15:42:55 +01:00
# Copyright (c) 2014-2020 The Dash Core developers
2014-10-23 03:48:19 +02:00
# Distributed under the MIT software license, see the accompanying
2014-07-08 18:07:23 +02:00
# file COPYING or http://www.opensource.org/licenses/mit-license.php.
2019-01-07 10:55:35 +01:00
""" Base class for RPC testing. """
2020-01-04 12:22:41 +01:00
import copy
2017-03-28 11:24:14 +02:00
from collections import deque
2017-06-02 11:32:55 +02:00
from enum import Enum
2016-05-10 18:27:31 +02:00
import logging
import optparse
2014-07-08 18:07:23 +02:00
import os
2017-08-11 18:09:51 +02:00
import pdb
2014-07-08 18:07:23 +02:00
import shutil
2017-05-07 15:13:29 +02:00
import sys
2014-07-08 18:07:23 +02:00
import tempfile
2017-03-22 13:03:26 +01:00
import time
2017-06-29 17:37:19 +02:00
import traceback
2019-01-23 17:36:51 +01:00
from concurrent . futures import ThreadPoolExecutor
2014-07-08 18:07:23 +02:00
2017-06-29 17:37:19 +02:00
from . authproxy import JSONRPCException
from . import coverage
2017-08-15 23:34:07 +02:00
from . test_node import TestNode
2015-10-11 07:41:19 +02:00
from . util import (
2017-05-07 15:13:29 +02:00
PortSeed ,
MAX_NODES ,
2019-06-20 18:37:09 +02:00
assert_equal ,
2017-05-07 15:13:29 +02:00
check_json_precision ,
2015-10-11 07:41:19 +02:00
connect_nodes_bi ,
2018-10-20 16:08:02 +02:00
connect_nodes ,
2019-06-20 18:37:09 +02:00
copy_datadir ,
2017-05-02 20:02:55 +02:00
disconnect_nodes ,
2019-10-09 18:48:12 +02:00
force_finish_mnsync ,
2018-03-14 14:18:44 +01:00
get_datadir_path ,
2017-05-07 15:13:29 +02:00
initialize_datadir ,
p2p_port ,
set_node_times ,
2019-06-20 18:37:09 +02:00
satoshi_round ,
2015-10-11 07:41:19 +02:00
sync_blocks ,
sync_mempools ,
2019-10-01 16:14:26 +02:00
wait_until ,
)
2014-07-08 18:07:23 +02:00
2017-06-02 11:32:55 +02:00
class TestStatus ( Enum ) :
PASSED = 1
FAILED = 2
SKIPPED = 3
TEST_EXIT_PASSED = 0
TEST_EXIT_FAILED = 1
TEST_EXIT_SKIPPED = 77
2019-08-09 01:14:11 +02:00
GENESISTIME = 1417713337
2017-10-17 21:03:09 +02:00
class BitcoinTestFramework ( ) :
2017-05-07 15:13:29 +02:00
""" Base class for a bitcoin test script.
2017-09-01 18:47:13 +02:00
Individual bitcoin test scripts should subclass this class and override the set_test_params ( ) and run_test ( ) methods .
Individual tests can also override the following methods to customize the test setup :
2017-05-07 15:13:29 +02:00
- add_options ( )
- setup_chain ( )
- setup_network ( )
2017-09-01 18:47:13 +02:00
- setup_nodes ( )
2017-05-07 15:13:29 +02:00
2017-09-01 18:47:13 +02:00
The __init__ ( ) and main ( ) methods should not be overridden .
2017-05-07 15:13:29 +02:00
This class also contains various public and private helper methods . """
2016-05-20 15:16:51 +02:00
def __init__ ( self ) :
2017-09-01 18:47:13 +02:00
""" Sets test framework defaults. Do not override this method. Instead, override the set_test_params() method """
2016-05-20 15:16:51 +02:00
self . setup_clean_chain = False
2017-06-29 17:37:19 +02:00
self . nodes = [ ]
self . mocktime = 0
2018-01-12 23:24:36 +01:00
self . supports_cli = False
2017-09-01 18:47:13 +02:00
self . set_test_params ( )
2016-05-20 15:16:51 +02:00
2017-09-01 18:47:13 +02:00
assert hasattr ( self , " num_nodes " ) , " Test must set self.num_nodes in set_test_params() "
2014-07-08 18:07:23 +02:00
def main ( self ) :
2017-09-01 18:47:13 +02:00
""" Main function. This should not be overridden by the subclass test scripts. """
2014-07-08 18:07:23 +02:00
parser = optparse . OptionParser ( usage = " % prog [options] " )
parser . add_option ( " --nocleanup " , dest = " nocleanup " , default = False , action = " store_true " ,
2016-03-04 08:25:16 +01:00
help = " Leave dashds and test.* datadir on exit or error " )
2015-04-23 14:19:00 +02:00
parser . add_option ( " --noshutdown " , dest = " noshutdown " , default = False , action = " store_true " ,
2016-03-04 08:25:16 +01:00
help = " Don ' t stop dashds after the test execution " )
2017-06-29 17:37:19 +02:00
parser . add_option ( " --srcdir " , dest = " srcdir " , default = os . path . normpath ( os . path . dirname ( os . path . realpath ( __file__ ) ) + " /../../../src " ) ,
2016-03-04 08:25:16 +01:00
help = " Source directory containing dashd/dash-cli (default: %d efault) " )
2017-06-29 17:37:19 +02:00
parser . add_option ( " --cachedir " , dest = " cachedir " , default = os . path . normpath ( os . path . dirname ( os . path . realpath ( __file__ ) ) + " /../../cache " ) ,
2016-08-17 12:12:55 +02:00
help = " Directory for caching pregenerated datadirs " )
2017-05-22 08:59:11 +02:00
parser . add_option ( " --tmpdir " , dest = " tmpdir " , help = " Root directory for datadirs " )
2017-03-09 21:16:20 +01:00
parser . add_option ( " -l " , " --loglevel " , dest = " loglevel " , default = " INFO " ,
help = " log events at this level and higher to the console. Can be set to DEBUG, INFO, WARNING, ERROR or CRITICAL. Passing --loglevel DEBUG will output all logs to console. Note that logs at all levels are always written to the test_framework.log file in the temporary test directory. " )
2014-10-23 19:11:20 +02:00
parser . add_option ( " --tracerpc " , dest = " trace_rpc " , default = False , action = " store_true " ,
help = " Print out all RPC calls as they are made " )
2016-05-10 18:27:31 +02:00
parser . add_option ( " --portseed " , dest = " port_seed " , default = os . getpid ( ) , type = ' int ' ,
help = " The seed to use for assigning port numbers (default: current process id) " )
2015-10-11 07:41:19 +02:00
parser . add_option ( " --coveragedir " , dest = " coveragedir " ,
help = " Write tested RPC commands into this directory " )
2017-05-03 08:37:14 +02:00
parser . add_option ( " --configfile " , dest = " configfile " ,
help = " Location of the test framework config file " )
2017-08-11 18:09:51 +02:00
parser . add_option ( " --pdbonfailure " , dest = " pdbonfailure " , default = False , action = " store_true " ,
help = " Attach a python debugger if test fails " )
2018-01-12 23:24:36 +01:00
parser . add_option ( " --usecli " , dest = " usecli " , default = False , action = " store_true " ,
help = " use bitcoin-cli instead of RPC for all commands " )
2014-07-08 18:07:23 +02:00
self . add_options ( parser )
( self . options , self . args ) = parser . parse_args ( )
2016-05-10 18:27:31 +02:00
PortSeed . n = self . options . port_seed
2017-06-29 17:37:19 +02:00
os . environ [ ' PATH ' ] = self . options . srcdir + " : " + self . options . srcdir + " /qt: " + os . environ [ ' PATH ' ]
2014-07-08 18:07:23 +02:00
check_json_precision ( )
2017-10-18 16:52:44 +02:00
self . options . cachedir = os . path . abspath ( self . options . cachedir )
2017-03-09 21:16:20 +01:00
# Set up temp directory and start logging
2017-05-22 08:59:11 +02:00
if self . options . tmpdir :
2017-10-18 16:52:44 +02:00
self . options . tmpdir = os . path . abspath ( self . options . tmpdir )
2017-05-22 08:59:11 +02:00
os . makedirs ( self . options . tmpdir , exist_ok = False )
else :
self . options . tmpdir = tempfile . mkdtemp ( prefix = " test " )
2017-03-09 21:16:20 +01:00
self . _start_logging ( )
2017-06-02 11:32:55 +02:00
success = TestStatus . FAILED
2017-03-09 21:16:20 +01:00
2014-07-08 18:07:23 +02:00
try :
2018-01-12 23:24:36 +01:00
if self . options . usecli and not self . supports_cli :
raise SkipTest ( " --usecli specified but test does not support using CLI " )
2014-10-20 14:14:04 +02:00
self . setup_chain ( )
self . setup_network ( )
self . run_test ( )
2017-06-02 11:32:55 +02:00
success = TestStatus . PASSED
2014-07-09 03:24:40 +02:00
except JSONRPCException as e :
2017-03-09 21:16:20 +01:00
self . log . exception ( " JSONRPC error " )
2017-06-02 11:32:55 +02:00
except SkipTest as e :
self . log . warning ( " Test Skipped: %s " % e . message )
success = TestStatus . SKIPPED
2014-07-08 18:07:23 +02:00
except AssertionError as e :
2017-03-09 21:16:20 +01:00
self . log . exception ( " Assertion failed " )
2016-04-14 12:14:24 +02:00
except KeyError as e :
2017-03-09 21:16:20 +01:00
self . log . exception ( " Key error " )
2014-07-08 18:07:23 +02:00
except Exception as e :
2017-03-09 21:16:20 +01:00
self . log . exception ( " Unexpected exception caught during testing " )
2016-05-09 17:01:55 +02:00
except KeyboardInterrupt as e :
2017-03-09 21:16:20 +01:00
self . log . warning ( " Exiting after keyboard interrupt " )
2014-07-08 18:07:23 +02:00
2017-08-11 18:09:51 +02:00
if success == TestStatus . FAILED and self . options . pdbonfailure :
print ( " Testcase failed. Attaching python debugger. Enter ? for help " )
pdb . set_trace ( )
2015-04-23 14:19:00 +02:00
if not self . options . noshutdown :
2017-03-09 21:16:20 +01:00
self . log . info ( " Stopping nodes " )
2019-02-21 19:37:16 +01:00
try :
2017-06-02 11:32:55 +02:00
if self . nodes :
self . stop_nodes ( )
2019-02-21 19:37:16 +01:00
except BaseException as e :
success = False
2019-03-08 09:05:00 +01:00
self . log . exception ( " Unexpected exception caught during shutdown " )
2015-04-23 14:19:00 +02:00
else :
2018-04-08 17:05:44 +02:00
for node in self . nodes :
node . cleanup_on_exit = False
2017-03-09 21:16:20 +01:00
self . log . info ( " Note: dashds were not stopped and may still be running " )
2015-04-20 11:50:33 +02:00
2017-06-02 11:32:55 +02:00
if not self . options . nocleanup and not self . options . noshutdown and success != TestStatus . FAILED :
2017-03-09 21:16:20 +01:00
self . log . info ( " Cleaning up " )
2014-07-08 18:07:23 +02:00
shutil . rmtree ( self . options . tmpdir )
2016-05-25 11:52:25 +02:00
else :
2017-03-09 21:16:20 +01:00
self . log . warning ( " Not cleaning up dir %s " % self . options . tmpdir )
2016-12-02 16:40:50 +01:00
if os . getenv ( " PYTHON_DEBUG " , " " ) :
# Dump the end of the debug logs, to aid in debugging rare
# travis failures.
import glob
2017-03-28 11:24:14 +02:00
filenames = [ self . options . tmpdir + " /test_framework.log " ]
filenames + = glob . glob ( self . options . tmpdir + " /node*/regtest/debug.log " )
2016-12-02 16:40:50 +01:00
MAX_LINES_TO_PRINT = 1000
2017-03-28 11:24:14 +02:00
for fn in filenames :
try :
with open ( fn , ' r ' ) as f :
2017-06-29 17:37:19 +02:00
print ( " From " , fn , " : " )
2017-03-28 11:24:14 +02:00
print ( " " . join ( deque ( f , MAX_LINES_TO_PRINT ) ) )
except OSError :
print ( " Opening file %s failed. " % fn )
traceback . print_exc ( )
2017-06-02 11:32:55 +02:00
if success == TestStatus . PASSED :
2017-03-09 21:16:20 +01:00
self . log . info ( " Tests successful " )
2017-06-02 11:32:55 +02:00
sys . exit ( TEST_EXIT_PASSED )
elif success == TestStatus . SKIPPED :
self . log . info ( " Test skipped " )
sys . exit ( TEST_EXIT_SKIPPED )
2014-07-08 18:07:23 +02:00
else :
2017-03-09 21:16:20 +01:00
self . log . error ( " Test failed. Test logging available at %s /test_framework.log " , self . options . tmpdir )
logging . shutdown ( )
2017-06-02 11:32:55 +02:00
sys . exit ( TEST_EXIT_FAILED )
2015-04-28 18:39:47 +02:00
2017-09-01 18:47:13 +02:00
# Methods to override in subclass test scripts.
def set_test_params ( self ) :
""" Tests must this method to change default values for number of nodes, topology, etc """
raise NotImplementedError
2017-05-07 15:13:29 +02:00
2017-09-01 18:47:13 +02:00
def add_options ( self , parser ) :
""" Override this method to add command-line options to the test """
pass
2017-06-29 17:37:19 +02:00
2017-09-01 18:47:13 +02:00
def setup_chain ( self ) :
""" Override this method to customize blockchain setup """
self . log . info ( " Initializing test directory " + self . options . tmpdir )
if self . setup_clean_chain :
self . _initialize_chain_clean ( )
self . set_genesis_mocktime ( )
else :
self . _initialize_chain ( )
self . set_cache_mocktime ( )
2017-06-29 17:37:19 +02:00
2017-09-01 18:47:13 +02:00
def setup_network ( self ) :
""" Override this method to customize test network topology """
self . setup_nodes ( )
# Connect the nodes as a "chain". This allows us
# to split the network between nodes 1 and 2 to get
# two halves that can work on competing chains.
for i in range ( self . num_nodes - 1 ) :
connect_nodes_bi ( self . nodes , i , i + 1 )
self . sync_all ( )
def setup_nodes ( self ) :
""" Override this method to customize test node setup """
extra_args = None
2019-09-24 00:54:26 +02:00
stderr = None
2017-09-01 18:47:13 +02:00
if hasattr ( self , " extra_args " ) :
extra_args = self . extra_args
2019-09-24 00:54:26 +02:00
if hasattr ( self , " stderr " ) :
stderr = self . stderr
self . add_nodes ( self . num_nodes , extra_args , stderr = stderr )
2017-09-01 18:47:13 +02:00
self . start_nodes ( )
def run_test ( self ) :
""" Tests must override this method to define test logic """
raise NotImplementedError
2017-06-29 17:37:19 +02:00
2017-09-01 18:47:13 +02:00
# Public helper methods. These can be accessed by the subclass test scripts.
2017-06-29 17:37:19 +02:00
2017-09-01 18:47:13 +02:00
def add_nodes ( self , num_nodes , extra_args = None , rpchost = None , timewait = None , binary = None , stderr = None ) :
""" Instantiate TestNode objects """
2017-05-07 15:13:29 +02:00
2017-06-29 17:37:19 +02:00
if extra_args is None :
2017-08-15 23:34:07 +02:00
extra_args = [ [ ] ] * num_nodes
2017-06-29 17:37:19 +02:00
if binary is None :
binary = [ None ] * num_nodes
assert_equal ( len ( extra_args ) , num_nodes )
assert_equal ( len ( binary ) , num_nodes )
2019-09-23 23:15:31 +02:00
old_num_nodes = len ( self . nodes )
2017-09-01 18:47:13 +02:00
for i in range ( num_nodes ) :
2018-01-12 23:24:36 +01:00
self . nodes . append ( TestNode ( old_num_nodes + i , self . options . tmpdir , extra_args [ i ] , rpchost , timewait = timewait , binary = binary [ i ] , stderr = stderr , mocktime = self . mocktime , coverage_dir = self . options . coveragedir , use_cli = self . options . usecli ) )
2017-09-01 18:47:13 +02:00
2018-01-19 17:44:27 +01:00
def start_node ( self , i , * args , * * kwargs ) :
2019-09-26 15:50:24 +02:00
""" Start a dashd """
2017-09-01 18:47:13 +02:00
node = self . nodes [ i ]
2018-01-19 17:44:27 +01:00
node . start ( * args , * * kwargs )
2017-09-01 18:47:13 +02:00
node . wait_for_rpc_connection ( )
if self . options . coveragedir is not None :
coverage . write_all_rpc_commands ( self . options . coveragedir , node . rpc )
2018-01-19 17:44:27 +01:00
def start_nodes ( self , extra_args = None , stderr = None , * args , * * kwargs ) :
2017-09-01 18:47:13 +02:00
""" Start multiple dashds """
if extra_args is None :
extra_args = [ None ] * self . num_nodes
assert_equal ( len ( extra_args ) , self . num_nodes )
2017-06-29 17:37:19 +02:00
try :
2017-09-01 18:47:13 +02:00
for i , node in enumerate ( self . nodes ) :
2018-01-19 17:44:27 +01:00
node . start ( extra_args [ i ] , stderr , * args , * * kwargs )
2017-09-01 18:47:13 +02:00
for node in self . nodes :
2017-08-15 23:34:07 +02:00
node . wait_for_rpc_connection ( )
2017-06-29 17:37:19 +02:00
except :
# If one node failed to start, stop the others
self . stop_nodes ( )
raise
2017-08-15 23:34:07 +02:00
if self . options . coveragedir is not None :
2017-09-01 18:47:13 +02:00
for node in self . nodes :
2017-08-15 23:34:07 +02:00
coverage . write_all_rpc_commands ( self . options . coveragedir , node . rpc )
Merge #14670: http: Fix HTTP server shutdown
28479f926f21f2a91bec5a06671c60e5b0c55532 qa: Test bitcond shutdown (João Barbosa)
8d3f46ec3938e2ba17654fecacd1d2629f9915fd http: Remove timeout to exit event loop (João Barbosa)
e98a9eede2fb48ff33a020acc888cbcd83e24bbf http: Remove unnecessary event_base_loopexit call (João Barbosa)
6b13580f4e3842c11abd9b8bee7255fb2472b6fe http: Unlisten sockets after all workers quit (João Barbosa)
18e968581697078c36a3c3818f8906cf134ccadd http: Send "Connection: close" header if shutdown is requested (João Barbosa)
02e1e4eff6cda0bfc24b455a7c1583394cbff6eb rpc: Add wait argument to stop (João Barbosa)
Pull request description:
Fixes #11777. Reverts #11006. Replaces #13501.
With this change the HTTP server will exit gracefully, meaning that all requests will finish processing and sending the response, even if this means to wait more than 2 seconds (current time allowed to exit the event loop).
Another small change is that connections are accepted even when the server is stopping, but HTTP requests are rejected. This can be improved later, especially if chunked replies are implemented.
Briefly, before this PR, this is the order or events when a request arrives (RPC `stop`):
1. `bufferevent_disable(..., EV_READ)`
2. `StartShutdown()`
3. `evhttp_del_accept_socket(...)`
4. `ThreadHTTP` terminates (event loop exits) because there are no active or pending events thanks to 1. and 3.
5. client doesn't get the response thanks to 4.
This can be verified by applying
```diff
// Event loop will exit after current HTTP requests have been handled, so
// this reply will get back to the client.
StartShutdown();
+ MilliSleep(2000);
return "Bitcoin server stopping";
}
```
and checking the log output:
```
Received a POST request for / from 127.0.0.1:62443
ThreadRPCServer method=stop user=__cookie__
Interrupting HTTP server
** Exited http event loop
Interrupting HTTP RPC server
Interrupting RPC
tor: Thread interrupt
Shutdown: In progress...
torcontrol thread exit
Stopping HTTP RPC server
addcon thread exit
opencon thread exit
Unregistering HTTP handler for / (exactmatch 1)
Unregistering HTTP handler for /wallet/ (exactmatch 0)
Stopping RPC
RPC stopped.
Stopping HTTP server
Waiting for HTTP worker threads to exit
msghand thread exit
net thread exit
... sleep 2 seconds ...
Waiting for HTTP event thread to exit
Stopped HTTP server
```
For this reason point 3. is moved right after all HTTP workers quit. In that moment HTTP replies are queued in the event loop which keeps spinning util all connections are closed. In order to trigger the server side close with keep alive connections (implicit in HTTP/1.1) the header `Connection: close` is sent if shutdown was requested. This can be tested by
```
bitcoind -regtest
nc localhost 18443
POST / HTTP/1.1
Authorization: Basic ...
Content-Type: application/json
Content-Length: 44
{"jsonrpc": "2.0","method":"stop","id":123}
```
Summing up, this PR:
- removes explicit event loop exit — event loop exits once there are no active or pending events
- changes the moment the listening sockets are removed — explained above
- sends header `Connection: close` on active requests when shutdown was requested which is relevant when it's a persistent connection (default in HTTP 1.1) — libevent is aware of this header and closes the connection gracefully
- removes event loop explicit break after 2 seconds timeout
Tree-SHA512: 4dac1e86abe388697c1e2dedbf31fb36a394cfafe5e64eadbf6ed01d829542785a8c3b91d1ab680d3f03f912d14fc87176428041141441d25dcb6c98a1e069d8
2018-12-06 17:42:52 +01:00
def stop_node ( self , i , wait = 0 ) :
2019-08-09 01:21:57 +02:00
""" Stop a dashd test node """
Merge #14670: http: Fix HTTP server shutdown
28479f926f21f2a91bec5a06671c60e5b0c55532 qa: Test bitcond shutdown (João Barbosa)
8d3f46ec3938e2ba17654fecacd1d2629f9915fd http: Remove timeout to exit event loop (João Barbosa)
e98a9eede2fb48ff33a020acc888cbcd83e24bbf http: Remove unnecessary event_base_loopexit call (João Barbosa)
6b13580f4e3842c11abd9b8bee7255fb2472b6fe http: Unlisten sockets after all workers quit (João Barbosa)
18e968581697078c36a3c3818f8906cf134ccadd http: Send "Connection: close" header if shutdown is requested (João Barbosa)
02e1e4eff6cda0bfc24b455a7c1583394cbff6eb rpc: Add wait argument to stop (João Barbosa)
Pull request description:
Fixes #11777. Reverts #11006. Replaces #13501.
With this change the HTTP server will exit gracefully, meaning that all requests will finish processing and sending the response, even if this means to wait more than 2 seconds (current time allowed to exit the event loop).
Another small change is that connections are accepted even when the server is stopping, but HTTP requests are rejected. This can be improved later, especially if chunked replies are implemented.
Briefly, before this PR, this is the order or events when a request arrives (RPC `stop`):
1. `bufferevent_disable(..., EV_READ)`
2. `StartShutdown()`
3. `evhttp_del_accept_socket(...)`
4. `ThreadHTTP` terminates (event loop exits) because there are no active or pending events thanks to 1. and 3.
5. client doesn't get the response thanks to 4.
This can be verified by applying
```diff
// Event loop will exit after current HTTP requests have been handled, so
// this reply will get back to the client.
StartShutdown();
+ MilliSleep(2000);
return "Bitcoin server stopping";
}
```
and checking the log output:
```
Received a POST request for / from 127.0.0.1:62443
ThreadRPCServer method=stop user=__cookie__
Interrupting HTTP server
** Exited http event loop
Interrupting HTTP RPC server
Interrupting RPC
tor: Thread interrupt
Shutdown: In progress...
torcontrol thread exit
Stopping HTTP RPC server
addcon thread exit
opencon thread exit
Unregistering HTTP handler for / (exactmatch 1)
Unregistering HTTP handler for /wallet/ (exactmatch 0)
Stopping RPC
RPC stopped.
Stopping HTTP server
Waiting for HTTP worker threads to exit
msghand thread exit
net thread exit
... sleep 2 seconds ...
Waiting for HTTP event thread to exit
Stopped HTTP server
```
For this reason point 3. is moved right after all HTTP workers quit. In that moment HTTP replies are queued in the event loop which keeps spinning util all connections are closed. In order to trigger the server side close with keep alive connections (implicit in HTTP/1.1) the header `Connection: close` is sent if shutdown was requested. This can be tested by
```
bitcoind -regtest
nc localhost 18443
POST / HTTP/1.1
Authorization: Basic ...
Content-Type: application/json
Content-Length: 44
{"jsonrpc": "2.0","method":"stop","id":123}
```
Summing up, this PR:
- removes explicit event loop exit — event loop exits once there are no active or pending events
- changes the moment the listening sockets are removed — explained above
- sends header `Connection: close` on active requests when shutdown was requested which is relevant when it's a persistent connection (default in HTTP 1.1) — libevent is aware of this header and closes the connection gracefully
- removes event loop explicit break after 2 seconds timeout
Tree-SHA512: 4dac1e86abe388697c1e2dedbf31fb36a394cfafe5e64eadbf6ed01d829542785a8c3b91d1ab680d3f03f912d14fc87176428041141441d25dcb6c98a1e069d8
2018-12-06 17:42:52 +01:00
self . nodes [ i ] . stop_node ( wait = wait )
2017-09-06 20:02:08 +02:00
self . nodes [ i ] . wait_until_stopped ( )
2019-08-09 01:19:22 +02:00
Merge #14670: http: Fix HTTP server shutdown
28479f926f21f2a91bec5a06671c60e5b0c55532 qa: Test bitcond shutdown (João Barbosa)
8d3f46ec3938e2ba17654fecacd1d2629f9915fd http: Remove timeout to exit event loop (João Barbosa)
e98a9eede2fb48ff33a020acc888cbcd83e24bbf http: Remove unnecessary event_base_loopexit call (João Barbosa)
6b13580f4e3842c11abd9b8bee7255fb2472b6fe http: Unlisten sockets after all workers quit (João Barbosa)
18e968581697078c36a3c3818f8906cf134ccadd http: Send "Connection: close" header if shutdown is requested (João Barbosa)
02e1e4eff6cda0bfc24b455a7c1583394cbff6eb rpc: Add wait argument to stop (João Barbosa)
Pull request description:
Fixes #11777. Reverts #11006. Replaces #13501.
With this change the HTTP server will exit gracefully, meaning that all requests will finish processing and sending the response, even if this means to wait more than 2 seconds (current time allowed to exit the event loop).
Another small change is that connections are accepted even when the server is stopping, but HTTP requests are rejected. This can be improved later, especially if chunked replies are implemented.
Briefly, before this PR, this is the order or events when a request arrives (RPC `stop`):
1. `bufferevent_disable(..., EV_READ)`
2. `StartShutdown()`
3. `evhttp_del_accept_socket(...)`
4. `ThreadHTTP` terminates (event loop exits) because there are no active or pending events thanks to 1. and 3.
5. client doesn't get the response thanks to 4.
This can be verified by applying
```diff
// Event loop will exit after current HTTP requests have been handled, so
// this reply will get back to the client.
StartShutdown();
+ MilliSleep(2000);
return "Bitcoin server stopping";
}
```
and checking the log output:
```
Received a POST request for / from 127.0.0.1:62443
ThreadRPCServer method=stop user=__cookie__
Interrupting HTTP server
** Exited http event loop
Interrupting HTTP RPC server
Interrupting RPC
tor: Thread interrupt
Shutdown: In progress...
torcontrol thread exit
Stopping HTTP RPC server
addcon thread exit
opencon thread exit
Unregistering HTTP handler for / (exactmatch 1)
Unregistering HTTP handler for /wallet/ (exactmatch 0)
Stopping RPC
RPC stopped.
Stopping HTTP server
Waiting for HTTP worker threads to exit
msghand thread exit
net thread exit
... sleep 2 seconds ...
Waiting for HTTP event thread to exit
Stopped HTTP server
```
For this reason point 3. is moved right after all HTTP workers quit. In that moment HTTP replies are queued in the event loop which keeps spinning util all connections are closed. In order to trigger the server side close with keep alive connections (implicit in HTTP/1.1) the header `Connection: close` is sent if shutdown was requested. This can be tested by
```
bitcoind -regtest
nc localhost 18443
POST / HTTP/1.1
Authorization: Basic ...
Content-Type: application/json
Content-Length: 44
{"jsonrpc": "2.0","method":"stop","id":123}
```
Summing up, this PR:
- removes explicit event loop exit — event loop exits once there are no active or pending events
- changes the moment the listening sockets are removed — explained above
- sends header `Connection: close` on active requests when shutdown was requested which is relevant when it's a persistent connection (default in HTTP 1.1) — libevent is aware of this header and closes the connection gracefully
- removes event loop explicit break after 2 seconds timeout
Tree-SHA512: 4dac1e86abe388697c1e2dedbf31fb36a394cfafe5e64eadbf6ed01d829542785a8c3b91d1ab680d3f03f912d14fc87176428041141441d25dcb6c98a1e069d8
2018-12-06 17:42:52 +01:00
def stop_nodes ( self , wait = 0 ) :
2019-08-09 01:21:57 +02:00
""" Stop multiple dashd test nodes """
2017-08-15 23:34:07 +02:00
for node in self . nodes :
# Issue RPC to stop nodes
Merge #14670: http: Fix HTTP server shutdown
28479f926f21f2a91bec5a06671c60e5b0c55532 qa: Test bitcond shutdown (João Barbosa)
8d3f46ec3938e2ba17654fecacd1d2629f9915fd http: Remove timeout to exit event loop (João Barbosa)
e98a9eede2fb48ff33a020acc888cbcd83e24bbf http: Remove unnecessary event_base_loopexit call (João Barbosa)
6b13580f4e3842c11abd9b8bee7255fb2472b6fe http: Unlisten sockets after all workers quit (João Barbosa)
18e968581697078c36a3c3818f8906cf134ccadd http: Send "Connection: close" header if shutdown is requested (João Barbosa)
02e1e4eff6cda0bfc24b455a7c1583394cbff6eb rpc: Add wait argument to stop (João Barbosa)
Pull request description:
Fixes #11777. Reverts #11006. Replaces #13501.
With this change the HTTP server will exit gracefully, meaning that all requests will finish processing and sending the response, even if this means to wait more than 2 seconds (current time allowed to exit the event loop).
Another small change is that connections are accepted even when the server is stopping, but HTTP requests are rejected. This can be improved later, especially if chunked replies are implemented.
Briefly, before this PR, this is the order or events when a request arrives (RPC `stop`):
1. `bufferevent_disable(..., EV_READ)`
2. `StartShutdown()`
3. `evhttp_del_accept_socket(...)`
4. `ThreadHTTP` terminates (event loop exits) because there are no active or pending events thanks to 1. and 3.
5. client doesn't get the response thanks to 4.
This can be verified by applying
```diff
// Event loop will exit after current HTTP requests have been handled, so
// this reply will get back to the client.
StartShutdown();
+ MilliSleep(2000);
return "Bitcoin server stopping";
}
```
and checking the log output:
```
Received a POST request for / from 127.0.0.1:62443
ThreadRPCServer method=stop user=__cookie__
Interrupting HTTP server
** Exited http event loop
Interrupting HTTP RPC server
Interrupting RPC
tor: Thread interrupt
Shutdown: In progress...
torcontrol thread exit
Stopping HTTP RPC server
addcon thread exit
opencon thread exit
Unregistering HTTP handler for / (exactmatch 1)
Unregistering HTTP handler for /wallet/ (exactmatch 0)
Stopping RPC
RPC stopped.
Stopping HTTP server
Waiting for HTTP worker threads to exit
msghand thread exit
net thread exit
... sleep 2 seconds ...
Waiting for HTTP event thread to exit
Stopped HTTP server
```
For this reason point 3. is moved right after all HTTP workers quit. In that moment HTTP replies are queued in the event loop which keeps spinning util all connections are closed. In order to trigger the server side close with keep alive connections (implicit in HTTP/1.1) the header `Connection: close` is sent if shutdown was requested. This can be tested by
```
bitcoind -regtest
nc localhost 18443
POST / HTTP/1.1
Authorization: Basic ...
Content-Type: application/json
Content-Length: 44
{"jsonrpc": "2.0","method":"stop","id":123}
```
Summing up, this PR:
- removes explicit event loop exit — event loop exits once there are no active or pending events
- changes the moment the listening sockets are removed — explained above
- sends header `Connection: close` on active requests when shutdown was requested which is relevant when it's a persistent connection (default in HTTP 1.1) — libevent is aware of this header and closes the connection gracefully
- removes event loop explicit break after 2 seconds timeout
Tree-SHA512: 4dac1e86abe388697c1e2dedbf31fb36a394cfafe5e64eadbf6ed01d829542785a8c3b91d1ab680d3f03f912d14fc87176428041141441d25dcb6c98a1e069d8
2018-12-06 17:42:52 +01:00
node . stop_node ( wait = wait )
2017-06-29 17:37:19 +02:00
2017-08-15 23:34:07 +02:00
for node in self . nodes :
# Wait for nodes to stop
2017-09-06 20:02:08 +02:00
node . wait_until_stopped ( )
2017-06-29 17:37:19 +02:00
2017-09-25 13:16:08 +02:00
def restart_node ( self , i , extra_args = None ) :
""" Stop and start a test node """
self . stop_node ( i )
self . start_node ( i , extra_args )
2018-01-19 17:44:27 +01:00
def assert_start_raises_init_error ( self , i , extra_args = None , expected_msg = None , * args , * * kwargs ) :
2017-06-29 17:37:19 +02:00
with tempfile . SpooledTemporaryFile ( max_size = 2 * * 16 ) as log_stderr :
try :
2018-01-19 17:44:27 +01:00
self . start_node ( i , extra_args , stderr = log_stderr , * args , * * kwargs )
2017-06-29 17:37:19 +02:00
self . stop_node ( i )
except Exception as e :
2019-08-09 01:21:57 +02:00
assert ' dashd exited ' in str ( e ) # node must have shutdown
2017-08-15 23:34:07 +02:00
self . nodes [ i ] . running = False
self . nodes [ i ] . process = None
2017-06-29 17:37:19 +02:00
if expected_msg is not None :
log_stderr . seek ( 0 )
stderr = log_stderr . read ( ) . decode ( ' utf-8 ' )
if expected_msg not in stderr :
raise AssertionError ( " Expected error \" " + expected_msg + " \" not found in: \n " + stderr )
else :
if expected_msg is None :
2019-08-09 01:21:57 +02:00
assert_msg = " dashd should have exited with an error "
2017-06-29 17:37:19 +02:00
else :
2019-08-09 01:21:57 +02:00
assert_msg = " dashd should have exited with expected error " + expected_msg
2017-06-29 17:37:19 +02:00
raise AssertionError ( assert_msg )
def wait_for_node_exit ( self , i , timeout ) :
2017-08-15 23:34:07 +02:00
self . nodes [ i ] . process . wait ( timeout )
2017-05-07 15:13:29 +02:00
def split_network ( self ) :
"""
Split the network of four nodes into nodes 0 / 1 and 2 / 3.
"""
disconnect_nodes ( self . nodes [ 1 ] , 2 )
disconnect_nodes ( self . nodes [ 2 ] , 1 )
self . sync_all ( [ self . nodes [ : 2 ] , self . nodes [ 2 : ] ] )
def join_network ( self ) :
"""
Join the ( previously split ) network halves together .
"""
connect_nodes_bi ( self . nodes , 1 , 2 )
self . sync_all ( )
def sync_all ( self , node_groups = None ) :
if not node_groups :
node_groups = [ self . nodes ]
for group in node_groups :
sync_blocks ( group )
sync_mempools ( group )
2019-08-09 01:14:11 +02:00
def disable_mocktime ( self ) :
self . mocktime = 0
2017-06-29 17:37:19 +02:00
2019-08-09 01:14:11 +02:00
def bump_mocktime ( self , t ) :
self . mocktime + = t
2017-06-29 17:37:19 +02:00
2019-08-09 01:14:11 +02:00
def set_cache_mocktime ( self ) :
# For backwared compatibility of the python scripts
# with previous versions of the cache, set MOCKTIME
# to regtest genesis time + (201 * 156)
self . mocktime = GENESISTIME + ( 201 * 156 )
2017-06-29 17:37:19 +02:00
2019-08-09 01:14:11 +02:00
def set_genesis_mocktime ( self ) :
self . mocktime = GENESISTIME
2017-06-29 17:37:19 +02:00
2017-05-07 15:13:29 +02:00
# Private helper methods. These should not be accessed by the subclass test scripts.
2019-02-26 23:04:56 +01:00
def _start_logging ( self ) :
# Add logger and logging handlers
self . log = logging . getLogger ( ' TestFramework ' )
self . log . setLevel ( logging . DEBUG )
# Create file handler to log all messages
fh = logging . FileHandler ( self . options . tmpdir + ' /test_framework.log ' )
fh . setLevel ( logging . DEBUG )
# Create console handler to log messages to stderr. By default this logs only error messages, but can be configured with --loglevel.
ch = logging . StreamHandler ( sys . stdout )
# User can provide log level as a number or string (eg DEBUG). loglevel was caught as a string, so try to convert it to an int
ll = int ( self . options . loglevel ) if self . options . loglevel . isdigit ( ) else self . options . loglevel . upper ( )
ch . setLevel ( ll )
2019-08-09 01:21:57 +02:00
# Format logs the same as dashd's debug.log with microprecision (so log files can be concatenated and sorted)
2017-06-29 17:37:19 +02:00
formatter = logging . Formatter ( fmt = ' %(asctime)s . %(msecs)03d 000 %(name)s ( %(levelname)s ): %(message)s ' , datefmt = ' % Y- % m- %d % H: % M: % S ' )
2017-03-22 13:03:26 +01:00
formatter . converter = time . gmtime
2019-02-26 23:04:56 +01:00
fh . setFormatter ( formatter )
ch . setFormatter ( formatter )
# add the handlers to the logger
self . log . addHandler ( fh )
self . log . addHandler ( ch )
if self . options . trace_rpc :
rpc_logger = logging . getLogger ( " BitcoinRPC " )
rpc_logger . setLevel ( logging . DEBUG )
rpc_handler = logging . StreamHandler ( sys . stdout )
rpc_handler . setLevel ( logging . DEBUG )
rpc_logger . addHandler ( rpc_handler )
2017-09-01 18:47:13 +02:00
def _initialize_chain ( self , extra_args = None , stderr = None ) :
2017-05-07 15:13:29 +02:00
""" Initialize a pre-mined blockchain for use by the test.
Create a cache of a 200 - block - long chain ( with wallet ) for MAX_NODES
Afterward , create num_nodes copies from the cache . """
2017-09-01 18:47:13 +02:00
assert self . num_nodes < = MAX_NODES
2017-05-07 15:13:29 +02:00
create_cache = False
for i in range ( MAX_NODES ) :
2018-03-14 14:18:44 +01:00
if not os . path . isdir ( get_datadir_path ( self . options . cachedir , i ) ) :
2017-05-07 15:13:29 +02:00
create_cache = True
break
if create_cache :
self . log . debug ( " Creating data directories from cached datadir " )
# find and delete old cache directories if any exist
for i in range ( MAX_NODES ) :
2018-03-14 14:18:44 +01:00
if os . path . isdir ( get_datadir_path ( self . options . cachedir , i ) ) :
shutil . rmtree ( get_datadir_path ( self . options . cachedir , i ) )
2017-05-07 15:13:29 +02:00
# Create cache directories, run dashds:
2019-08-09 01:14:11 +02:00
self . set_genesis_mocktime ( )
2017-05-07 15:13:29 +02:00
for i in range ( MAX_NODES ) :
2017-09-01 18:47:13 +02:00
datadir = initialize_datadir ( self . options . cachedir , i )
2019-06-19 20:01:39 +02:00
args = [ os . getenv ( " DASHD " , " dashd " ) , " -server " , " -keypool=1 " , " -datadir= " + datadir , " -discover=0 " , " -mocktime= " + str ( GENESISTIME ) ]
2017-05-07 15:13:29 +02:00
if i > 0 :
args . append ( " -connect=127.0.0.1: " + str ( p2p_port ( 0 ) ) )
2019-06-20 18:36:17 +02:00
if extra_args is not None :
args . extend ( extra_args )
2017-09-01 18:47:13 +02:00
self . nodes . append ( TestNode ( i , self . options . cachedir , extra_args = [ ] , rpchost = None , timewait = None , binary = None , stderr = stderr , mocktime = self . mocktime , coverage_dir = None ) )
2017-08-15 23:34:07 +02:00
self . nodes [ i ] . args = args
2017-09-01 18:47:13 +02:00
self . start_node ( i )
2017-05-07 15:13:29 +02:00
2017-08-15 23:34:07 +02:00
# Wait for RPC connections to be ready
for node in self . nodes :
node . wait_for_rpc_connection ( )
2017-05-07 15:13:29 +02:00
# Create a 200-block-long chain; each of the 4 first nodes
# gets 25 mature blocks and 25 immature.
# Note: To preserve compatibility with older versions of
# initialize_chain, only 4 nodes will generate coins.
#
# blocks are created with timestamps 10 minutes apart
# starting from 2010 minutes in the past
2019-06-19 20:01:39 +02:00
block_time = GENESISTIME
2017-05-07 15:13:29 +02:00
for i in range ( 2 ) :
for peer in range ( 4 ) :
for j in range ( 25 ) :
set_node_times ( self . nodes , block_time )
self . nodes [ peer ] . generate ( 1 )
2019-06-19 20:01:39 +02:00
block_time + = 156
2017-05-07 15:13:29 +02:00
# Must sync before next peer starts generating blocks
sync_blocks ( self . nodes )
# Shut them down, and clean up cache directories:
self . stop_nodes ( )
self . nodes = [ ]
2017-06-29 17:37:19 +02:00
self . disable_mocktime ( )
2018-03-14 14:18:44 +01:00
def cache_path ( n , * paths ) :
return os . path . join ( get_datadir_path ( self . options . cachedir , n ) , " regtest " , * paths )
2017-05-07 15:13:29 +02:00
for i in range ( MAX_NODES ) :
2018-03-14 14:18:44 +01:00
for entry in os . listdir ( cache_path ( i ) ) :
2020-03-24 17:42:29 +01:00
if entry not in [ ' wallets ' , ' chainstate ' , ' blocks ' , ' evodb ' , ' llmq ' , ' backups ' ] :
2018-03-14 14:18:44 +01:00
os . remove ( cache_path ( i , entry ) )
2017-09-01 18:47:13 +02:00
for i in range ( self . num_nodes ) :
2018-03-14 14:18:44 +01:00
from_dir = get_datadir_path ( self . options . cachedir , i )
to_dir = get_datadir_path ( self . options . tmpdir , i )
2017-05-07 15:13:29 +02:00
shutil . copytree ( from_dir , to_dir )
2017-09-01 18:47:13 +02:00
initialize_datadir ( self . options . tmpdir , i ) # Overwrite port/rpcport in dash.conf
2017-05-07 15:13:29 +02:00
2017-09-01 18:47:13 +02:00
def _initialize_chain_clean ( self ) :
2017-05-07 15:13:29 +02:00
""" Initialize empty blockchain for use by the test.
Create an empty blockchain and num_nodes wallets .
Useful if a test case wants complete control over initialization . """
2017-09-01 18:47:13 +02:00
for i in range ( self . num_nodes ) :
initialize_datadir ( self . options . tmpdir , i )
2017-05-07 15:13:29 +02:00
2018-10-20 16:08:02 +02:00
MASTERNODE_COLLATERAL = 1000
class MasternodeInfo :
2018-12-31 08:12:36 +01:00
def __init__ ( self , proTxHash , ownerAddr , votingAddr , pubKeyOperator , keyOperator , collateral_address , collateral_txid , collateral_vout ) :
self . proTxHash = proTxHash
self . ownerAddr = ownerAddr
self . votingAddr = votingAddr
self . pubKeyOperator = pubKeyOperator
self . keyOperator = keyOperator
self . collateral_address = collateral_address
self . collateral_txid = collateral_txid
self . collateral_vout = collateral_vout
2018-10-20 16:08:02 +02:00
class DashTestFramework ( BitcoinTestFramework ) :
2019-12-05 19:28:33 +01:00
def set_dash_test_params ( self , num_nodes , masterodes_count , extra_args = None , fast_dip3_enforcement = False ) :
2018-10-20 16:08:02 +02:00
self . mn_count = masterodes_count
self . num_nodes = num_nodes
self . mninfo = [ ]
self . setup_clean_chain = True
self . is_network_split = False
# additional args
2019-12-05 19:28:33 +01:00
if extra_args is None :
extra_args = [ [ ] ] * num_nodes
assert_equal ( len ( extra_args ) , num_nodes )
2020-01-04 12:22:41 +01:00
self . extra_args = [ copy . deepcopy ( a ) for a in extra_args ]
2019-12-05 19:28:33 +01:00
self . extra_args [ 0 ] + = [ " -sporkkey=cP4EKFyJsHT39LDqgdcB43Y3YXjNyjb5Fuas1GQSeAtjnZWmZEQK " ]
2019-01-29 15:54:38 +01:00
self . fast_dip3_enforcement = fast_dip3_enforcement
if fast_dip3_enforcement :
2019-12-05 19:28:33 +01:00
for i in range ( 0 , num_nodes ) :
2020-01-04 12:22:41 +01:00
self . extra_args [ i ] . append ( " -dip3params=30:50 " )
2019-01-23 17:36:51 +01:00
2020-01-07 13:49:51 +01:00
# LLMQ default test params (no need to pass -llmqtestparams)
self . llmq_size = 3
self . llmq_threshold = 2
2020-01-04 12:21:16 +01:00
def set_dash_dip8_activation ( self , activate_after_block ) :
window = int ( ( activate_after_block + 2 ) / 3 )
threshold = int ( ( window + 1 ) / 2 )
for i in range ( 0 , self . num_nodes ) :
self . extra_args [ i ] . append ( " -vbparams=dip0008:0:999999999999: %d : %d " % ( window , threshold ) )
2020-01-07 13:49:51 +01:00
def set_dash_llmq_test_params ( self , llmq_size , llmq_threshold ) :
self . llmq_size = llmq_size
self . llmq_threshold = llmq_threshold
for i in range ( 0 , self . num_nodes ) :
self . extra_args [ i ] . append ( " -llmqtestparams= %d : %d " % ( self . llmq_size , self . llmq_threshold ) )
2018-10-20 16:08:02 +02:00
def create_simple_node ( self ) :
idx = len ( self . nodes )
2019-12-05 19:28:33 +01:00
self . add_nodes ( 1 , extra_args = [ self . extra_args [ idx ] ] )
2019-09-24 00:57:30 +02:00
self . start_node ( idx )
2018-10-20 16:08:02 +02:00
for i in range ( 0 , idx ) :
connect_nodes ( self . nodes [ i ] , idx )
def prepare_masternodes ( self ) :
2019-10-03 12:45:18 +02:00
self . log . info ( " Preparing %d masternodes " % self . mn_count )
2018-10-20 16:08:02 +02:00
for idx in range ( 0 , self . mn_count ) :
2019-04-04 08:08:44 +02:00
self . prepare_masternode ( idx )
def prepare_masternode ( self , idx ) :
bls = self . nodes [ 0 ] . bls ( ' generate ' )
address = self . nodes [ 0 ] . getnewaddress ( )
txid = self . nodes [ 0 ] . sendtoaddress ( address , MASTERNODE_COLLATERAL )
txraw = self . nodes [ 0 ] . getrawtransaction ( txid , True )
collateral_vout = 0
for vout_idx in range ( 0 , len ( txraw [ " vout " ] ) ) :
vout = txraw [ " vout " ] [ vout_idx ]
if vout [ " value " ] == MASTERNODE_COLLATERAL :
collateral_vout = vout_idx
self . nodes [ 0 ] . lockunspent ( False , [ { ' txid ' : txid , ' vout ' : collateral_vout } ] )
# send to same address to reserve some funds for fees
self . nodes [ 0 ] . sendtoaddress ( address , 0.001 )
ownerAddr = self . nodes [ 0 ] . getnewaddress ( )
votingAddr = self . nodes [ 0 ] . getnewaddress ( )
rewardsAddr = self . nodes [ 0 ] . getnewaddress ( )
port = p2p_port ( len ( self . nodes ) + idx )
if ( idx % 2 ) == 0 :
self . nodes [ 0 ] . lockunspent ( True , [ { ' txid ' : txid , ' vout ' : collateral_vout } ] )
proTxHash = self . nodes [ 0 ] . protx ( ' register_fund ' , address , ' 127.0.0.1: %d ' % port , ownerAddr , bls [ ' public ' ] , votingAddr , 0 , rewardsAddr , address )
else :
2018-12-31 08:12:36 +01:00
self . nodes [ 0 ] . generate ( 1 )
2019-04-04 08:08:44 +02:00
proTxHash = self . nodes [ 0 ] . protx ( ' register ' , txid , collateral_vout , ' 127.0.0.1: %d ' % port , ownerAddr , bls [ ' public ' ] , votingAddr , 0 , rewardsAddr , address )
self . nodes [ 0 ] . generate ( 1 )
self . mninfo . append ( MasternodeInfo ( proTxHash , ownerAddr , votingAddr , bls [ ' public ' ] , bls [ ' secret ' ] , address , txid , collateral_vout ) )
self . sync_all ( )
2018-10-20 16:08:02 +02:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Prepared masternode %d : collateral_txid= %s , collateral_vout= %d , protxHash= %s " % ( idx , txid , collateral_vout , proTxHash ) )
2019-04-04 08:08:44 +02:00
def remove_mastermode ( self , idx ) :
mn = self . mninfo [ idx ]
rawtx = self . nodes [ 0 ] . createrawtransaction ( [ { " txid " : mn . collateral_txid , " vout " : mn . collateral_vout } ] , { self . nodes [ 0 ] . getnewaddress ( ) : 999.9999 } )
rawtx = self . nodes [ 0 ] . signrawtransaction ( rawtx )
self . nodes [ 0 ] . sendrawtransaction ( rawtx [ " hex " ] )
self . nodes [ 0 ] . generate ( 1 )
2018-12-31 08:12:36 +01:00
self . sync_all ( )
2019-04-04 08:08:44 +02:00
self . mninfo . remove ( mn )
2018-10-20 16:08:02 +02:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Removed masternode %d " , idx )
2019-01-23 17:36:51 +01:00
def prepare_datadirs ( self ) :
# stop faucet node so that we can copy the datadir
2019-07-04 16:48:01 +02:00
self . stop_node ( 0 )
2019-01-23 17:36:51 +01:00
start_idx = len ( self . nodes )
for idx in range ( 0 , self . mn_count ) :
copy_datadir ( 0 , idx + start_idx , self . options . tmpdir )
# restart faucet node
2019-09-24 00:57:30 +02:00
self . start_node ( 0 )
2019-01-23 17:36:51 +01:00
2018-12-31 08:12:36 +01:00
def start_masternodes ( self ) :
2019-10-03 12:45:18 +02:00
self . log . info ( " Starting %d masternodes " , self . mn_count )
2018-12-31 08:12:36 +01:00
start_idx = len ( self . nodes )
2019-01-23 17:36:51 +01:00
2019-09-24 00:57:30 +02:00
self . add_nodes ( self . mn_count )
2019-01-23 17:36:51 +01:00
executor = ThreadPoolExecutor ( max_workers = 20 )
def do_start ( idx ) :
2019-12-05 19:28:33 +01:00
args = [ ' -masternodeblsprivkey= %s ' % self . mninfo [ idx ] . keyOperator ] + self . extra_args [ idx + start_idx ]
2019-09-24 00:57:30 +02:00
self . start_node ( idx + start_idx , extra_args = args )
2019-03-11 09:38:58 +01:00
self . mninfo [ idx ] . nodeIdx = idx + start_idx
2019-09-24 00:57:30 +02:00
self . mninfo [ idx ] . node = self . nodes [ idx + start_idx ]
2019-10-09 18:48:12 +02:00
force_finish_mnsync ( self . mninfo [ idx ] . node )
2019-01-23 17:36:51 +01:00
def do_connect ( idx ) :
2019-12-05 19:31:55 +01:00
# Connect to the control node only, masternodes should take care of intra-quorum connections themselves
connect_nodes ( self . mninfo [ idx ] . node , 0 )
2019-01-23 17:36:51 +01:00
jobs = [ ]
# start up nodes in parallel
for idx in range ( 0 , self . mn_count ) :
jobs . append ( executor . submit ( do_start , idx ) )
# wait for all nodes to start up
for job in jobs :
job . result ( )
jobs . clear ( )
# connect nodes in parallel
for idx in range ( 0 , self . mn_count ) :
jobs . append ( executor . submit ( do_connect , idx ) )
# wait for all nodes to connect
for job in jobs :
job . result ( )
jobs . clear ( )
executor . shutdown ( )
2018-10-20 16:08:02 +02:00
def setup_network ( self ) :
2019-10-03 12:45:18 +02:00
self . log . info ( " Creating and starting controller node " )
2019-12-05 19:28:33 +01:00
self . add_nodes ( 1 , extra_args = [ self . extra_args [ 0 ] ] )
2019-09-24 00:57:30 +02:00
self . start_node ( 0 )
2018-10-20 16:08:02 +02:00
required_balance = MASTERNODE_COLLATERAL * self . mn_count + 1
2019-10-03 12:45:18 +02:00
self . log . info ( " Generating %d coins " % required_balance )
2018-10-20 16:08:02 +02:00
while self . nodes [ 0 ] . getbalance ( ) < required_balance :
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
set_node_times ( self . nodes , self . mocktime )
2020-01-04 12:21:16 +01:00
self . nodes [ 0 ] . generate ( 10 )
2019-10-03 12:45:18 +02:00
num_simple_nodes = self . num_nodes - self . mn_count - 1
self . log . info ( " Creating and starting %s simple nodes " , num_simple_nodes )
for i in range ( 0 , num_simple_nodes ) :
2018-10-20 16:08:02 +02:00
self . create_simple_node ( )
2018-12-31 08:12:36 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Activating DIP3 " )
2019-01-29 15:54:38 +01:00
if not self . fast_dip3_enforcement :
2019-01-23 17:36:51 +01:00
while self . nodes [ 0 ] . getblockcount ( ) < 500 :
self . nodes [ 0 ] . generate ( 10 )
2018-12-31 08:12:36 +01:00
self . sync_all ( )
# create masternodes
self . prepare_masternodes ( )
2019-01-23 17:36:51 +01:00
self . prepare_datadirs ( )
2018-12-31 08:12:36 +01:00
self . start_masternodes ( )
2019-12-05 19:31:55 +01:00
# non-masternodes where disconnected from the control node during prepare_datadirs,
# let's reconnect them back to make sure they receive updates
for i in range ( 0 , num_simple_nodes ) :
connect_nodes ( self . nodes [ i + 1 ] , 0 )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
set_node_times ( self . nodes , self . mocktime )
2018-10-20 16:08:02 +02:00
self . nodes [ 0 ] . generate ( 1 )
# sync nodes
self . sync_all ( )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
set_node_times ( self . nodes , self . mocktime )
2018-12-31 08:12:36 +01:00
2018-10-20 16:08:02 +02:00
mn_info = self . nodes [ 0 ] . masternodelist ( " status " )
assert ( len ( mn_info ) == self . mn_count )
for status in mn_info . values ( ) :
assert ( status == ' ENABLED ' )
2019-03-18 14:50:44 +01:00
def create_raw_tx ( self , node_from , node_to , amount , min_inputs , max_inputs ) :
2018-10-20 16:08:02 +02:00
assert ( min_inputs < = max_inputs )
# fill inputs
inputs = [ ]
balances = node_from . listunspent ( )
in_amount = 0.0
last_amount = 0.0
for tx in balances :
if len ( inputs ) < min_inputs :
input = { }
input [ " txid " ] = tx [ ' txid ' ]
input [ ' vout ' ] = tx [ ' vout ' ]
in_amount + = float ( tx [ ' amount ' ] )
inputs . append ( input )
elif in_amount > amount :
break
elif len ( inputs ) < max_inputs :
input = { }
input [ " txid " ] = tx [ ' txid ' ]
input [ ' vout ' ] = tx [ ' vout ' ]
in_amount + = float ( tx [ ' amount ' ] )
inputs . append ( input )
else :
input = { }
input [ " txid " ] = tx [ ' txid ' ]
input [ ' vout ' ] = tx [ ' vout ' ]
in_amount - = last_amount
in_amount + = float ( tx [ ' amount ' ] )
inputs [ - 1 ] = input
last_amount = float ( tx [ ' amount ' ] )
2019-03-18 14:50:44 +01:00
assert ( len ( inputs ) > = min_inputs )
assert ( len ( inputs ) < = max_inputs )
assert ( in_amount > = amount )
2018-10-20 16:08:02 +02:00
# fill outputs
receiver_address = node_to . getnewaddress ( )
change_address = node_from . getnewaddress ( )
fee = 0.001
outputs = { }
outputs [ receiver_address ] = satoshi_round ( amount )
outputs [ change_address ] = satoshi_round ( in_amount - amount - fee )
rawtx = node_from . createrawtransaction ( inputs , outputs )
2019-05-07 14:14:33 +02:00
ret = node_from . signrawtransaction ( rawtx )
decoded = node_from . decoderawtransaction ( ret [ ' hex ' ] )
ret = { * * decoded , * * ret }
return ret
2018-10-20 16:08:02 +02:00
2019-12-06 10:05:58 +01:00
def wait_for_tx ( self , txid , node , expected = True , timeout = 15 ) :
def check_tx ( ) :
try :
return node . getrawtransaction ( txid )
except :
return False
if wait_until ( check_tx , timeout = timeout , sleep = 0.5 , do_assert = expected ) and not expected :
raise AssertionError ( " waiting unexpectedly succeeded " )
def wait_for_instantlock ( self , txid , node , expected = True , timeout = 15 ) :
2019-10-01 16:14:26 +02:00
def check_instantlock ( ) :
2019-03-19 08:38:16 +01:00
try :
2019-10-01 16:14:26 +02:00
return node . getrawtransaction ( txid , True ) [ " instantlock " ]
2019-03-19 08:38:16 +01:00
except :
2019-10-01 16:14:26 +02:00
return False
2019-12-06 10:05:58 +01:00
if wait_until ( check_instantlock , timeout = timeout , sleep = 0.5 , do_assert = expected ) and not expected :
raise AssertionError ( " waiting unexpectedly succeeded " )
2018-10-20 16:08:02 +02:00
2019-12-06 10:05:58 +01:00
def wait_for_chainlocked_block ( self , node , block_hash , expected = True , timeout = 15 ) :
2019-10-02 02:11:10 +02:00
def check_chainlocked_block ( ) :
try :
block = node . getblock ( block_hash )
return block [ " confirmations " ] > 0 and block [ " chainlock " ]
except :
2019-01-18 11:51:31 +01:00
return False
2019-12-06 10:05:58 +01:00
if wait_until ( check_chainlocked_block , timeout = timeout , sleep = 0.1 , do_assert = expected ) and not expected :
raise AssertionError ( " waiting unexpectedly succeeded " )
2019-10-02 02:11:10 +02:00
2019-10-02 15:24:57 +02:00
def wait_for_chainlocked_block_all_nodes ( self , block_hash , timeout = 15 ) :
2019-10-02 02:11:10 +02:00
for node in self . nodes :
2019-12-06 10:05:58 +01:00
self . wait_for_chainlocked_block ( node , block_hash , timeout = timeout )
2019-10-02 02:11:10 +02:00
2019-10-02 15:24:57 +02:00
def wait_for_best_chainlock ( self , node , block_hash , timeout = 15 ) :
wait_until ( lambda : node . getbestchainlock ( ) [ " blockhash " ] == block_hash , timeout = timeout , sleep = 0.1 )
2019-10-02 02:11:10 +02:00
def wait_for_sporks_same ( self , timeout = 30 ) :
def check_sporks_same ( ) :
sporks = self . nodes [ 0 ] . spork ( ' show ' )
return all ( node . spork ( ' show ' ) == sporks for node in self . nodes [ 1 : ] )
wait_until ( check_sporks_same , timeout = timeout , sleep = 0.5 )
2019-01-18 11:51:31 +01:00
2020-03-30 10:37:49 +02:00
def wait_for_quorum_connections ( self , expected_connections , nodes , timeout = 60 , wait_proc = None ) :
2020-01-08 08:01:26 +01:00
def check_quorum_connections ( ) :
all_ok = True
2020-03-30 10:37:49 +02:00
for node in nodes :
2020-01-08 08:01:26 +01:00
s = node . quorum ( " dkgstatus " )
if s [ " session " ] == { } :
continue
if " quorumConnections " not in s :
all_ok = False
break
s = s [ " quorumConnections " ]
if " llmq_test " not in s :
all_ok = False
break
cnt = 0
for c in s [ " llmq_test " ] :
if c [ " connected " ] :
cnt + = 1
if cnt < expected_connections :
all_ok = False
break
2020-03-27 22:59:02 +01:00
if not all_ok and wait_proc is not None :
wait_proc ( )
2020-01-08 08:01:26 +01:00
return all_ok
2020-03-31 07:06:41 +02:00
wait_until ( check_quorum_connections , timeout = timeout , sleep = 1 )
2020-01-08 08:01:26 +01:00
2020-03-30 14:51:26 +02:00
def wait_for_masternode_probes ( self , expected_probes , mninfos , timeout = 30 , wait_proc = None ) :
def check_probes ( ) :
for mn in mninfos :
l = mn . node . protx ( ' list ' , ' registered ' , 1 )
cnt = 0
for mn2 in l :
if mn2 [ ' proTxHash ' ] != mn . proTxHash :
if mn2 [ ' metaInfo ' ] [ ' lastOutboundSuccessElapsed ' ] < = 60 :
cnt + = 1
if cnt < expected_probes :
if wait_proc is not None :
wait_proc ( )
return False
return True
wait_until ( check_probes , timeout = timeout , sleep = 1 )
2020-03-30 10:37:49 +02:00
def wait_for_quorum_phase ( self , quorum_hash , phase , expected_member_count , check_received_messages , check_received_messages_count , mninfos , timeout = 30 , sleep = 0.1 ) :
2019-10-02 02:11:10 +02:00
def check_dkg_session ( ) :
2019-01-18 11:51:31 +01:00
all_ok = True
2020-01-07 13:49:51 +01:00
member_count = 0
2020-03-30 10:37:49 +02:00
for mn in mninfos :
2019-02-01 08:49:01 +01:00
s = mn . node . quorum ( " dkgstatus " ) [ " session " ]
2020-01-07 13:49:51 +01:00
if " llmq_test " not in s :
continue
member_count + = 1
s = s [ " llmq_test " ]
2020-03-26 13:25:24 +01:00
if s [ " quorumHash " ] != quorum_hash :
all_ok = False
break
2019-01-18 11:51:31 +01:00
if " phase " not in s :
all_ok = False
break
if s [ " phase " ] != phase :
all_ok = False
break
if check_received_messages is not None :
2019-01-23 15:02:56 +01:00
if s [ check_received_messages ] < check_received_messages_count :
2019-01-18 11:51:31 +01:00
all_ok = False
break
2020-01-07 13:49:51 +01:00
if all_ok and member_count != expected_member_count :
2020-03-27 22:59:02 +01:00
return False
2019-10-02 02:11:10 +02:00
return all_ok
2020-03-27 15:12:03 +01:00
wait_until ( check_dkg_session , timeout = timeout , sleep = sleep )
2019-01-18 11:51:31 +01:00
2020-03-30 10:37:49 +02:00
def wait_for_quorum_commitment ( self , quorum_hash , nodes , timeout = 15 ) :
2019-10-02 02:11:10 +02:00
def check_dkg_comitments ( ) :
2019-01-18 11:51:31 +01:00
all_ok = True
2020-03-30 10:37:49 +02:00
for node in nodes :
2019-02-01 08:49:01 +01:00
s = node . quorum ( " dkgstatus " )
if " minableCommitments " not in s :
all_ok = False
break
s = s [ " minableCommitments " ]
2020-01-07 13:49:51 +01:00
if " llmq_test " not in s :
2019-01-18 11:51:31 +01:00
all_ok = False
break
2020-03-26 13:25:24 +01:00
s = s [ " llmq_test " ]
if s [ " quorumHash " ] != quorum_hash :
all_ok = False
break
2019-10-02 02:11:10 +02:00
return all_ok
wait_until ( check_dkg_comitments , timeout = timeout , sleep = 0.1 )
2019-01-18 11:51:31 +01:00
2020-03-30 14:51:26 +02:00
def mine_quorum ( self , expected_members = None , expected_connections = 2 , expected_probes = 0 , expected_contributions = None , expected_complaints = 0 , expected_justifications = 0 , expected_commitments = None , mninfos = None ) :
2020-01-07 13:49:51 +01:00
if expected_members is None :
expected_members = self . llmq_size
if expected_contributions is None :
expected_contributions = self . llmq_size
if expected_commitments is None :
expected_commitments = self . llmq_size
2020-03-30 10:37:49 +02:00
if mninfos is None :
mninfos = self . mninfo
2020-01-07 13:49:51 +01:00
2020-03-30 14:51:26 +02:00
self . log . info ( " Mining quorum: expected_members= %d , expected_connections= %d , expected_probes= %d , expected_contributions= %d , expected_complaints= %d , expected_justifications= %d , "
" expected_commitments= %d " % ( expected_members , expected_connections , expected_probes , expected_contributions , expected_complaints ,
2019-10-03 12:45:18 +02:00
expected_justifications , expected_commitments ) )
2020-03-30 10:37:49 +02:00
nodes = [ self . nodes [ 0 ] ] + [ mn . node for mn in mninfos ]
2019-01-18 11:51:31 +01:00
quorums = self . nodes [ 0 ] . quorum ( " list " )
# move forward to next DKG
skip_count = 24 - ( self . nodes [ 0 ] . getblockcount ( ) % 24 )
if skip_count != 0 :
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( skip_count )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2019-01-18 11:51:31 +01:00
2020-03-26 13:25:24 +01:00
q = self . nodes [ 0 ] . getbestblockhash ( )
2019-10-03 12:45:18 +02:00
self . log . info ( " Waiting for phase 1 (init) " )
2020-03-27 15:12:03 +01:00
def bump_time ( ) :
2020-03-31 07:06:41 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
self . wait_for_quorum_phase ( q , 1 , expected_members , None , 0 , mninfos )
self . wait_for_quorum_connections ( expected_connections , nodes , wait_proc = bump_time )
2020-03-30 14:51:26 +02:00
self . wait_for_masternode_probes ( expected_probes , mninfos , wait_proc = bump_time )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( 2 )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2019-01-18 11:51:31 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Waiting for phase 2 (contribute) " )
2020-03-30 10:37:49 +02:00
self . wait_for_quorum_phase ( q , 2 , expected_members , " receivedContributions " , expected_contributions , mninfos )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( 2 )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2019-01-18 11:51:31 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Waiting for phase 3 (complain) " )
2020-03-30 10:37:49 +02:00
self . wait_for_quorum_phase ( q , 3 , expected_members , " receivedComplaints " , expected_complaints , mninfos )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( 2 )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2019-01-18 11:51:31 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Waiting for phase 4 (justify) " )
2020-03-30 10:37:49 +02:00
self . wait_for_quorum_phase ( q , 4 , expected_members , " receivedJustifications " , expected_justifications , mninfos )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( 2 )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2019-01-18 11:51:31 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Waiting for phase 5 (commit) " )
2020-03-30 10:37:49 +02:00
self . wait_for_quorum_phase ( q , 5 , expected_members , " receivedPrematureCommitments " , expected_commitments , mninfos )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( 2 )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2019-01-18 11:51:31 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Waiting for phase 6 (mining) " )
2020-03-30 10:37:49 +02:00
self . wait_for_quorum_phase ( q , 6 , expected_members , None , 0 , mninfos )
2019-01-18 11:51:31 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Waiting final commitment " )
2020-03-30 10:37:49 +02:00
self . wait_for_quorum_commitment ( q , nodes )
2019-01-18 11:51:31 +01:00
2019-10-03 12:45:18 +02:00
self . log . info ( " Mining final commitment " )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( 1 )
while quorums == self . nodes [ 0 ] . quorum ( " list " ) :
2019-05-21 10:01:57 +02:00
time . sleep ( 2 )
2019-08-09 01:14:11 +02:00
self . bump_mocktime ( 1 )
2020-03-30 10:37:49 +02:00
set_node_times ( nodes , self . mocktime )
2019-01-18 11:51:31 +01:00
self . nodes [ 0 ] . generate ( 1 )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2020-01-07 13:49:51 +01:00
new_quorum = self . nodes [ 0 ] . quorum ( " list " , 1 ) [ " llmq_test " ] [ 0 ]
2019-10-03 12:45:18 +02:00
quorum_info = self . nodes [ 0 ] . quorum ( " info " , 100 , new_quorum )
2019-01-18 11:51:31 +01:00
2019-04-04 08:15:31 +02:00
# Mine 8 (SIGN_HEIGHT_OFFSET) more blocks to make sure that the new quorum gets eligable for signing sessions
self . nodes [ 0 ] . generate ( 8 )
2020-03-30 10:37:49 +02:00
sync_blocks ( nodes )
2018-10-20 16:08:02 +02:00
2019-10-03 12:45:18 +02:00
self . log . info ( " New quorum: height= %d , quorumHash= %s , minedBlock= %s " % ( quorum_info [ " height " ] , new_quorum , quorum_info [ " minedBlock " ] ) )
2019-04-04 08:15:57 +02:00
return new_quorum
2020-03-26 08:42:54 +01:00
def get_quorum_masternodes ( self , q ) :
qi = self . nodes [ 0 ] . quorum ( ' info ' , 100 , q )
result = [ ]
for m in qi [ ' members ' ] :
for mn in self . mninfo :
if mn . proTxHash == m [ ' proTxHash ' ] :
result . append ( mn )
break
return result
2019-12-06 10:05:58 +01:00
def wait_for_mnauth ( self , node , count , timeout = 10 ) :
def test ( ) :
pi = node . getpeerinfo ( )
c = 0
for p in pi :
if " verified_proregtx_hash " in p and p [ " verified_proregtx_hash " ] != " " :
c + = 1
return c > = count
wait_until ( test , timeout = timeout )
2017-06-29 17:37:19 +02:00
class SkipTest ( Exception ) :
""" This exception is raised to skip a test """
def __init__ ( self , message ) :
self . message = message