2016-01-27 12:05:25 +01:00
// Copyright (c) 2015 The Bitcoin Core developers
// Distributed under the MIT software license, see the accompanying
// file COPYING or http://www.opensource.org/licenses/mit-license.php.
2020-03-19 23:46:56 +01:00
# include <httprpc.h>
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
2020-03-19 23:46:56 +01:00
# include <chainparams.h>
2022-03-24 05:13:51 +01:00
# include <crypto/hmac_sha256.h>
2020-03-19 23:46:56 +01:00
# include <httpserver.h>
# include <rpc/protocol.h>
# include <rpc/server.h>
2022-03-24 05:13:51 +01:00
# include <util/strencodings.h>
# include <util/system.h>
# include <util/translation.h>
2021-08-05 16:41:46 +02:00
# include <walletinitinterface.h>
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
# include <algorithm>
# include <iterator>
# include <map>
2018-04-02 20:31:40 +02:00
# include <memory>
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
# include <set>
# include <string>
2018-04-02 20:31:40 +02:00
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
# include <boost/algorithm/string.hpp> // boost::trim
2016-02-05 10:45:50 +01:00
/** WWW-Authenticate to present with 401 Unauthorized response */
static const char * WWW_AUTH_HEADER_DATA = " Basic realm= \" jsonrpc \" " ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
/** Simple one-shot callback timer to be used by the RPC mechanism to e.g.
2017-01-29 16:05:06 +01:00
* re - lock the wallet .
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
*/
class HTTPRPCTimer : public RPCTimerBase
{
public :
2018-09-20 23:57:13 +02:00
HTTPRPCTimer ( struct event_base * eventBase , std : : function < void ( ) > & func , int64_t millis ) :
2015-08-28 16:46:20 +02:00
ev ( eventBase , false , func )
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
2015-08-28 16:46:20 +02:00
struct timeval tv ;
tv . tv_sec = millis / 1000 ;
tv . tv_usec = ( millis % 1000 ) * 1000 ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
ev . trigger ( & tv ) ;
}
private :
HTTPEvent ev ;
} ;
class HTTPRPCTimerInterface : public RPCTimerInterface
{
public :
2017-08-17 22:59:56 +02:00
explicit HTTPRPCTimerInterface ( struct event_base * _base ) : base ( _base )
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
}
2018-02-15 08:29:15 +01:00
const char * Name ( ) override
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
return " HTTP " ;
}
2018-09-20 23:57:13 +02:00
RPCTimerBase * NewTimer ( std : : function < void ( ) > & func , int64_t millis ) override
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
2015-08-28 16:46:20 +02:00
return new HTTPRPCTimer ( base , func , millis ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
}
private :
struct event_base * base ;
} ;
/* Pre-base64-encoded authentication token */
static std : : string strRPCUserColonPass ;
/* Stored RPC timer interface (for unregistration) */
2017-11-09 21:22:08 +01:00
static std : : unique_ptr < HTTPRPCTimerInterface > httpRPCTimerInterface ;
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
/* RPC Auth Whitelist */
static std : : map < std : : string , std : : set < std : : string > > g_rpc_whitelist ;
static bool g_rpc_whitelist_default = false ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
static void JSONErrorReply ( HTTPRequest * req , const UniValue & objError , const UniValue & id )
{
// Send error reply from json-rpc error object
int nStatus = HTTP_INTERNAL_SERVER_ERROR ;
int code = find_value ( objError , " code " ) . get_int ( ) ;
if ( code = = RPC_INVALID_REQUEST )
nStatus = HTTP_BAD_REQUEST ;
else if ( code = = RPC_METHOD_NOT_FOUND )
nStatus = HTTP_NOT_FOUND ;
2020-11-24 02:39:50 +01:00
else if ( code = = RPC_PLATFORM_RESTRICTION ) {
nStatus = HTTP_FORBIDDEN ;
}
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
std : : string strReply = JSONRPCReply ( NullUniValue , objError , id ) ;
req - > WriteHeader ( " Content-Type " , " application/json " ) ;
req - > WriteReply ( nStatus , strReply ) ;
}
2015-11-11 16:49:32 +01:00
//This function checks username and password against -rpcauth
//entries from config file.
static bool multiUserAuthorized ( std : : string strUserPass )
2020-07-29 03:23:12 +02:00
{
2018-01-29 13:51:03 +01:00
if ( strUserPass . find ( ' : ' ) = = std : : string : : npos ) {
2015-11-11 16:49:32 +01:00
return false ;
}
2018-01-29 13:51:03 +01:00
std : : string strUser = strUserPass . substr ( 0 , strUserPass . find ( ' : ' ) ) ;
std : : string strPass = strUserPass . substr ( strUserPass . find ( ' : ' ) + 1 ) ;
2015-11-11 16:49:32 +01:00
2017-06-27 14:22:54 +02:00
for ( const std : : string & strRPCAuth : gArgs . GetArgs ( " -rpcauth " ) ) {
2015-11-11 16:49:32 +01:00
//Search for multi-user login/pass "rpcauth" from config
2017-06-27 14:22:54 +02:00
std : : vector < std : : string > vFields ;
boost : : split ( vFields , strRPCAuth , boost : : is_any_of ( " :$ " ) ) ;
if ( vFields . size ( ) ! = 3 ) {
//Incorrect formatting in config file
continue ;
}
std : : string strName = vFields [ 0 ] ;
if ( ! TimingResistantEqual ( strName , strUser ) ) {
continue ;
}
std : : string strSalt = vFields [ 1 ] ;
std : : string strHash = vFields [ 2 ] ;
static const unsigned int KEY_SIZE = 32 ;
unsigned char out [ KEY_SIZE ] ;
2019-10-28 13:41:45 +01:00
CHMAC_SHA256 ( reinterpret_cast < const unsigned char * > ( strSalt . data ( ) ) , strSalt . size ( ) ) . Write ( reinterpret_cast < const unsigned char * > ( strPass . data ( ) ) , strPass . size ( ) ) . Finalize ( out ) ;
2017-06-27 14:22:54 +02:00
std : : vector < unsigned char > hexvec ( out , out + KEY_SIZE ) ;
std : : string strHashFromPass = HexStr ( hexvec ) ;
if ( TimingResistantEqual ( strHashFromPass , strHash ) ) {
return true ;
2015-11-11 16:49:32 +01:00
}
}
return false ;
}
2016-10-19 15:01:33 +02:00
static bool RPCAuthorized ( const std : : string & strAuth , std : : string & strAuthUsernameOut )
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
if ( strRPCUserColonPass . empty ( ) ) // Belt-and-suspenders measure if InitRPCAuthentication was not called
return false ;
if ( strAuth . substr ( 0 , 6 ) ! = " Basic " )
return false ;
std : : string strUserPass64 = strAuth . substr ( 6 ) ;
boost : : trim ( strUserPass64 ) ;
std : : string strUserPass = DecodeBase64 ( strUserPass64 ) ;
2016-10-19 15:01:33 +02:00
2018-01-29 13:51:03 +01:00
if ( strUserPass . find ( ' : ' ) ! = std : : string : : npos )
strAuthUsernameOut = strUserPass . substr ( 0 , strUserPass . find ( ' : ' ) ) ;
2016-10-19 15:01:33 +02:00
2015-11-11 16:49:32 +01:00
//Check if authorized under single-user field
if ( TimingResistantEqual ( strUserPass , strRPCUserColonPass ) ) {
return true ;
}
return multiUserAuthorized ( strUserPass ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
}
2022-04-23 09:46:44 +02:00
static bool HTTPReq_JSONRPC ( const util : : Ref & context , HTTPRequest * req )
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
// JSONRPC handles only POST
if ( req - > GetRequestMethod ( ) ! = HTTPRequest : : POST ) {
req - > WriteReply ( HTTP_BAD_METHOD , " JSONRPC server handles only POST requests " ) ;
return false ;
}
// Check authorization
std : : pair < bool , std : : string > authHeader = req - > GetHeader ( " authorization " ) ;
if ( ! authHeader . first ) {
2016-02-05 10:45:50 +01:00
req - > WriteHeader ( " WWW-Authenticate " , WWW_AUTH_HEADER_DATA ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
req - > WriteReply ( HTTP_UNAUTHORIZED ) ;
return false ;
}
2022-04-23 09:46:44 +02:00
JSONRPCRequest jreq ( context ) ;
2018-03-27 20:43:34 +02:00
jreq . peerAddr = req - > GetPeer ( ) . ToString ( ) ;
2016-10-19 15:01:33 +02:00
if ( ! RPCAuthorized ( authHeader . second , jreq . authUser ) ) {
2018-03-27 20:43:34 +02:00
LogPrintf ( " ThreadRPCServer incorrect password attempt from %s \n " , jreq . peerAddr ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
/* Deter brute-forcing
If this results in a DoS the user really
shouldn ' t have their RPC port exposed . */
2021-07-13 11:31:17 +02:00
UninterruptibleSleep ( std : : chrono : : milliseconds { 250 } ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
2016-02-05 10:45:50 +01:00
req - > WriteHeader ( " WWW-Authenticate " , WWW_AUTH_HEADER_DATA ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
req - > WriteReply ( HTTP_UNAUTHORIZED ) ;
return false ;
}
try {
// Parse request
UniValue valRequest ;
if ( ! valRequest . read ( req - > ReadBody ( ) ) )
throw JSONRPCError ( RPC_PARSE_ERROR , " Parse error " ) ;
2016-10-19 15:01:33 +02:00
// Set the URI
jreq . URI = req - > GetURI ( ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
std : : string strReply ;
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
bool user_has_whitelist = g_rpc_whitelist . count ( jreq . authUser ) ;
if ( ! user_has_whitelist & & g_rpc_whitelist_default ) {
LogPrintf ( " RPC User %s not allowed to call any methods \n " , jreq . authUser ) ;
req - > WriteReply ( HTTP_FORBIDDEN ) ;
return false ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
// singleton request
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
} else if ( valRequest . isObject ( ) ) {
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
jreq . parse ( valRequest ) ;
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
if ( user_has_whitelist & & ! g_rpc_whitelist [ jreq . authUser ] . count ( jreq . strMethod ) ) {
LogPrintf ( " RPC User %s not allowed to call method %s \n " , jreq . authUser , jreq . strMethod ) ;
req - > WriteReply ( HTTP_FORBIDDEN ) ;
return false ;
}
2016-10-19 15:01:33 +02:00
UniValue result = tableRPC . execute ( jreq ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
// Send reply
strReply = JSONRPCReply ( result , NullUniValue , jreq . id ) ;
// array of requests
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
} else if ( valRequest . isArray ( ) ) {
if ( user_has_whitelist ) {
for ( unsigned int reqIdx = 0 ; reqIdx < valRequest . size ( ) ; reqIdx + + ) {
if ( ! valRequest [ reqIdx ] . isObject ( ) ) {
throw JSONRPCError ( RPC_INVALID_REQUEST , " Invalid Request object " ) ;
} else {
const UniValue & request = valRequest [ reqIdx ] . get_obj ( ) ;
// Parse method
std : : string strMethod = find_value ( request , " method " ) . get_str ( ) ;
if ( ! g_rpc_whitelist [ jreq . authUser ] . count ( strMethod ) ) {
LogPrintf ( " RPC User %s not allowed to call method %s \n " , jreq . authUser , strMethod ) ;
req - > WriteReply ( HTTP_FORBIDDEN ) ;
return false ;
}
}
}
}
2017-10-12 14:53:33 +02:00
strReply = JSONRPCExecBatch ( jreq , valRequest . get_array ( ) ) ;
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
}
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
else
throw JSONRPCError ( RPC_PARSE_ERROR , " Top-level object parse error " ) ;
req - > WriteHeader ( " Content-Type " , " application/json " ) ;
req - > WriteReply ( HTTP_OK , strReply ) ;
} catch ( const UniValue & objError ) {
JSONErrorReply ( req , objError , jreq . id ) ;
return false ;
} catch ( const std : : exception & e ) {
JSONErrorReply ( req , JSONRPCError ( RPC_PARSE_ERROR , e . what ( ) ) , jreq . id ) ;
return false ;
}
return true ;
}
static bool InitRPCAuthentication ( )
{
2019-06-24 18:44:27 +02:00
if ( gArgs . GetArg ( " -rpcpassword " , " " ) = = " " )
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
2020-05-19 09:21:53 +02:00
LogPrintf ( " Using random cookie authentication. \n " ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
if ( ! GenerateAuthCookie ( & strRPCUserColonPass ) ) {
return false ;
}
} else {
2020-07-19 08:15:38 +02:00
LogPrintf ( " Config options rpcuser and rpcpassword will soon be deprecated. Locally-run instances may remove rpcuser to use cookie-based auth, or may be replaced with rpcauth. Please see share/rpcauth for rpcauth auth generation. \n " ) ;
2019-06-24 18:44:27 +02:00
strRPCUserColonPass = gArgs . GetArg ( " -rpcuser " , " " ) + " : " + gArgs . GetArg ( " -rpcpassword " , " " ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
}
2018-07-10 16:18:40 +02:00
if ( gArgs . GetArg ( " -rpcauth " , " " ) ! = " " )
{
LogPrintf ( " Using rpcauth authentication. \n " ) ;
}
Merge #12763: Add RPC Whitelist Feature from #12248
2081442c421cc4376e5d7839f68fbe7630e89103 test: Add test for rpc_whitelist (Emil Engler)
7414d3820c833566b4f48c6c120a18bf53978c55 Add RPC Whitelist Feature from #12248 (Jeremy Rubin)
Pull request description:
Summary
====
This patch adds the RPC whitelisting feature requested in #12248. RPC Whitelists help enforce application policies for services being built on top of Bitcoin Core (e.g., your Lightning Node maybe shouldn't be adding new peers). The aim of this PR is not to make it advisable to connect your Bitcoin node to arbitrary services, but to reduce risk and prevent unintended access.
Using RPC Whitelists
====
The way it works is you specify (in your bitcoin.conf) configurations such as
```
rpcauth=user1:4cc74397d6e9972e5ee7671fd241$11849357f26a5be7809c68a032bc2b16ab5dcf6348ef3ed1cf30dae47b8bcc71
rpcauth=user2:181b4a25317bff60f3749adee7d6bca0$d9c331474f1322975fa170a2ffbcb176ba11644211746b27c1d317f265dd4ada
rpcauth=user3:a6c8a511b53b1edcf69c36984985e$13cfba0e626db19061c9d61fa58e712d0319c11db97ad845fa84517f454f6675
rpcwhitelist=user1:getnetworkinfo
rpcwhitelist=user2:getnetworkinfo,getwalletinfo, getbestblockhash
rpcwhitelistdefault=0
```
Now user1 can only call getnetworkinfo, user2 can only call getnetworkinfo or getwalletinfo, while user3 can still call all RPCs.
If any rpcwhitelist is set, act as if all users are subject to whitelists unless rpcwhitelistdefault is set to 0. If rpcwhitelistdefault is set to 1 and no rpcwhitelist is set, act as if all users are subject to whitelists.
Review Request
=====
In addition to normal review, would love specific review from someone working on LN (e.g., @ roasbeef) and someone working on an infrastructure team at an exchange (e.g., @ jimpo) to check that this works well with their system.
Notes
=====
The rpc list is spelling sensitive -- whitespace is stripped though. Spelling errors fail towards the RPC call being blocked, which is safer.
It was unclear to me if HTTPReq_JSONRPC is the best function to patch this functionality into, or if it would be better to place it in exec or somewhere else.
It was also unclear to me if it would be preferred to cache the whitelists on startup or parse them on every RPC as is done with multiUserAuthorized. I opted for the cached approach as I thought it was a bit cleaner.
Future Work
=====
In a future PR, I would like to add an inheritance scheme. This seemed more controversial so I didn't want to include that here. Inheritance semantics are tricky, but it would also make these whitelists easier to read.
It also might be good to add a `getrpcwhitelist` command to facilitate permission discovery.
Tests
=====
Thanks to @ emilengler for adding tests for this feature. The tests cover all cases except for where `rpcwhitelistdefault=1` is used, given difficulties around testing with the current test framework.
ACKs for top commit:
laanwj:
ACK 2081442c421cc4376e5d7839f68fbe7630e89103
Tree-SHA512: 0dc1ac6a6f2f4b0be9c9054d495dd17752fe7b3589aeab2c6ac4e1f91cf4e7e355deedcb5d76d707cbb5a949c2f989c871b74d6bf129351f429569a701adbcbf
2019-12-13 11:25:39 +01:00
g_rpc_whitelist_default = gArgs . GetBoolArg ( " -rpcwhitelistdefault " , gArgs . IsArgSet ( " -rpcwhitelist " ) ) ;
for ( const std : : string & strRPCWhitelist : gArgs . GetArgs ( " -rpcwhitelist " ) ) {
auto pos = strRPCWhitelist . find ( ' : ' ) ;
std : : string strUser = strRPCWhitelist . substr ( 0 , pos ) ;
bool intersect = g_rpc_whitelist . count ( strUser ) ;
std : : set < std : : string > & whitelist = g_rpc_whitelist [ strUser ] ;
if ( pos ! = std : : string : : npos ) {
std : : string strWhitelist = strRPCWhitelist . substr ( pos + 1 ) ;
std : : set < std : : string > new_whitelist ;
boost : : split ( new_whitelist , strWhitelist , boost : : is_any_of ( " , " ) ) ;
if ( intersect ) {
std : : set < std : : string > tmp_whitelist ;
std : : set_intersection ( new_whitelist . begin ( ) , new_whitelist . end ( ) ,
whitelist . begin ( ) , whitelist . end ( ) , std : : inserter ( tmp_whitelist , tmp_whitelist . end ( ) ) ) ;
new_whitelist = std : : move ( tmp_whitelist ) ;
}
whitelist = std : : move ( new_whitelist ) ;
}
}
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
return true ;
}
2022-04-23 09:46:44 +02:00
bool StartHTTPRPC ( const util : : Ref & context )
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
{
2019-05-22 23:51:39 +02:00
LogPrint ( BCLog : : RPC , " Starting HTTP RPC server \n " ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
if ( ! InitRPCAuthentication ( ) )
return false ;
2022-04-23 09:46:44 +02:00
auto handle_rpc = [ & context ] ( HTTPRequest * req , const std : : string & ) { return HTTPReq_JSONRPC ( context , req ) ; } ;
RegisterHTTPHandler ( " / " , true , handle_rpc ) ;
2021-08-05 16:41:46 +02:00
if ( g_wallet_init_interface . HasWalletSupport ( ) ) {
2022-04-23 09:46:44 +02:00
RegisterHTTPHandler ( " /wallet/ " , false , handle_rpc ) ;
2021-08-05 16:41:46 +02:00
}
2018-08-13 16:02:09 +02:00
struct event_base * eventBase = EventBase ( ) ;
assert ( eventBase ) ;
httpRPCTimerInterface = MakeUnique < HTTPRPCTimerInterface > ( eventBase ) ;
2017-11-09 21:22:08 +01:00
RPCSetTimerInterface ( httpRPCTimerInterface . get ( ) ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
return true ;
}
void InterruptHTTPRPC ( )
{
2019-05-22 23:51:39 +02:00
LogPrint ( BCLog : : RPC , " Interrupting HTTP RPC server \n " ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
}
void StopHTTPRPC ( )
{
2019-05-22 23:51:39 +02:00
LogPrint ( BCLog : : RPC , " Stopping HTTP RPC server \n " ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
UnregisterHTTPHandler ( " / " , true ) ;
2021-08-05 16:41:46 +02:00
if ( g_wallet_init_interface . HasWalletSupport ( ) ) {
UnregisterHTTPHandler ( " /wallet/ " , false ) ;
}
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
if ( httpRPCTimerInterface ) {
2017-11-09 21:22:08 +01:00
RPCUnsetTimerInterface ( httpRPCTimerInterface . get ( ) ) ;
httpRPCTimerInterface . reset ( ) ;
evhttpd implementation
- *Replace usage of boost::asio with [libevent2](http://libevent.org/)*.
boost::asio is not part of C++11, so unlike other boost there is no
forwards-compatibility reason to stick with it. Together with #4738 (convert
json_spirit to UniValue), this rids Bitcoin Core of the worst offenders with
regard to compile-time slowness.
- *Replace spit-and-duct-tape http server with evhttp*. Front-end http handling
is handled by libevent, a work queue (with configurable depth and parallelism)
is used to handle application requests.
- *Wrap HTTP request in C++ class*; this makes the application code mostly
HTTP-server-neutral
- *Refactor RPC to move all http-specific code to a separate file*.
Theoreticaly this can allow building without HTTP server but with another RPC
backend, e.g. Qt's debug console (currently not implemented) or future RPC
mechanisms people may want to use.
- *HTTP dispatch mechanism*; services (e.g., RPC, REST) register which URL
paths they want to handle.
By using a proven, high-performance asynchronous networking library (also used
by Tor) and HTTP server, problems such as #5674, #5655, #344 should be avoided.
What works? bitcoind, bitcoin-cli, bitcoin-qt. Unit tests and RPC/REST tests
pass. The aim for now is everything but SSL support.
Configuration options:
- `-rpcthreads`: repurposed as "number of work handler threads". Still
defaults to 4.
- `-rpcworkqueue`: maximum depth of work queue. When this is reached, new
requests will return a 500 Internal Error.
- `-rpctimeout`: inactivity time, in seconds, after which to disconnect a
client.
- `-debug=http`: low-level http activity logging
2015-01-23 07:53:17 +01:00
}
}