-
Notifications
You must be signed in to change notification settings - Fork 648
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
elasticsearch history api #1682 #1725
Changes from 5 commits
10a382f
00496a9
2a6d7b2
1737d45
6dda9fe
26e9f16
4fac171
29bf975
de76301
5838a38
1f6ac14
63f7aff
a7fc6b0
13d8034
07de45e
f1bedc7
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -26,7 +26,6 @@ | |
#include <graphene/chain/impacted.hpp> | ||
#include <graphene/chain/account_evaluator.hpp> | ||
#include <curl/curl.h> | ||
#include <graphene/utilities/elasticsearch.hpp> | ||
|
||
namespace graphene { namespace elasticsearch { | ||
|
||
|
@@ -59,6 +58,8 @@ class elasticsearch_plugin_impl | |
std::string _elasticsearch_index_prefix = "bitshares-"; | ||
bool _elasticsearch_operation_object = false; | ||
uint32_t _elasticsearch_start_es_after_block = 0; | ||
bool _elasticsearch_operation_string = true; | ||
std::string _elasticsearch_mode = "only_save"; | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. For more efficient comparison, make this an enum type. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. It is a bit tricky to do this because the Implemented at de76301 There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. changed the exception introduced here to |
||
CURL *curl; // curl handler | ||
vector <string> bulk_lines; // vector of op lines | ||
vector<std::string> prepare; | ||
|
@@ -223,9 +224,8 @@ void elasticsearch_plugin_impl::doOperationHistory(const optional <operation_his | |
adaptor_struct adaptor; | ||
os.op_object = adaptor.adapt(os.op_object.get_object()); | ||
} | ||
else | ||
if(_elasticsearch_operation_string) | ||
os.op = fc::json::to_string(oho->op); | ||
|
||
} | ||
|
||
void elasticsearch_plugin_impl::doBlock(uint32_t trx_in_block, const signed_block& b) | ||
|
@@ -436,17 +436,14 @@ void elasticsearch_plugin::plugin_set_program_options( | |
("elasticsearch-index-prefix", boost::program_options::value<std::string>(), "Add a prefix to the index(bitshares-)") | ||
("elasticsearch-operation-object", boost::program_options::value<bool>(), "Save operation as object(false)") | ||
("elasticsearch-start-es-after-block", boost::program_options::value<uint32_t>(), "Start doing ES job after block(0)") | ||
("elasticsearch-operation-string", boost::program_options::value<bool>(), "Save operation as string. Needed to serve history api calls(true)") | ||
("elasticsearch-mode", boost::program_options::value<std::string>(), "Mode of operation: only_save, only_query, all(only_save)") | ||
oxarbitrage marked this conversation as resolved.
Show resolved
Hide resolved
|
||
; | ||
cfg.add(cli); | ||
} | ||
|
||
void elasticsearch_plugin::plugin_initialize(const boost::program_options::variables_map& options) | ||
{ | ||
database().applied_block.connect( [&]( const signed_block& b) { | ||
if (!my->update_account_histories(b)) | ||
FC_THROW_EXCEPTION(graphene::chain::plugin_exception, "Error populating ES database, we are going to keep trying."); | ||
} ); | ||
|
||
my->_oho_index = database().add_index< primary_index< operation_history_index > >(); | ||
database().add_index< primary_index< account_transaction_history_index > >(); | ||
|
||
|
@@ -473,7 +470,23 @@ void elasticsearch_plugin::plugin_initialize(const boost::program_options::varia | |
} | ||
if (options.count("elasticsearch-start-es-after-block")) { | ||
my->_elasticsearch_start_es_after_block = options["elasticsearch-start-es-after-block"].as<uint32_t>(); | ||
} | ||
} | ||
if (options.count("elasticsearch-operation-string")) { | ||
my->_elasticsearch_operation_string = options["elasticsearch-operation-string"].as<bool>(); | ||
} | ||
if (options.count("elasticsearch-mode")) { | ||
my->_elasticsearch_mode = options["elasticsearch-mode"].as<std::string>(); | ||
} | ||
|
||
if(my->_elasticsearch_mode != "only_query") { | ||
if (my->_elasticsearch_mode == "all") | ||
my->_elasticsearch_operation_string = true; | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Silently overriding this will cause confusion. There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Agree, i am now not allowing the mode to be all if operation_string is false so no silent changes are made. |
||
|
||
database().applied_block.connect([&](const signed_block &b) { | ||
oxarbitrage marked this conversation as resolved.
Show resolved
Hide resolved
|
||
if (!my->update_account_histories(b)) | ||
FC_THROW_EXCEPTION(graphene::chain::plugin_exception, "Error populating ES database, we are going to keep trying."); | ||
}); | ||
} | ||
} | ||
|
||
void elasticsearch_plugin::plugin_startup() | ||
|
@@ -488,4 +501,139 @@ void elasticsearch_plugin::plugin_startup() | |
ilog("elasticsearch ACCOUNT HISTORY: plugin_startup() begin"); | ||
} | ||
|
||
operation_history_object elasticsearch_plugin::get_operation_by_id(operation_history_id_type id) | ||
{ | ||
const string operation_id_string = std::string(object_id_type(id)); | ||
|
||
const string query = R"( | ||
{ | ||
"query": { | ||
"bool": { | ||
"must": [ | ||
{ | ||
"query_string": { | ||
"query": "account_history.operation_id: )" + operation_id_string + R"(" | ||
} | ||
}, | ||
{ | ||
"range": { | ||
"block_data.block_time": { | ||
"gte": "now-20y", | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Put a TODO in your calendar for the year 2034: bump this. ;-) Can't you leave out this part of the query? There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I agree that the range could be omitted here. Also I am wondering why you do query_string and not
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. my apologies, this was not a production ready query but just a proof of concept probably pasted from kibana. i had changed this at 1f6ac14 There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. for the second query i removed the range but query_string needs to stay as we are making a lucene query there and not just matching a field. 63f7aff |
||
"lte": "now" | ||
} | ||
} | ||
} | ||
] | ||
} | ||
} | ||
} | ||
)"; | ||
|
||
auto es = prepareHistoryQuery(query); | ||
const auto response = graphene::utilities::simpleQuery(es); | ||
variant variant_response = fc::json::from_string(response); | ||
const auto source = variant_response["hits"]["hits"][size_t(0)]["_source"]; | ||
return fromEStoOperation(source); | ||
} | ||
|
||
vector<operation_history_object> elasticsearch_plugin::get_account_history( | ||
const account_id_type account_id, | ||
operation_history_id_type stop = operation_history_id_type(), | ||
unsigned limit = 100, | ||
operation_history_id_type start = operation_history_id_type()) | ||
{ | ||
const string account_id_string = std::string(object_id_type(account_id)); | ||
|
||
const auto stop_number = stop.instance.value; | ||
const auto start_number = start.instance.value; | ||
|
||
string range = ""; | ||
if(stop_number == 0) | ||
range = " AND operation_id_num: ["+fc::to_string(stop_number)+" TO "+fc::to_string(start_number)+"]"; | ||
else if(stop_number > 0) | ||
range = " AND operation_id_num: {"+fc::to_string(stop_number)+" TO "+fc::to_string(start_number)+"]"; | ||
|
||
const string query = R"( | ||
{ | ||
"size": )" + fc::to_string(limit) + R"(, | ||
"sort" : [{ "operation_id_num" : {"order" : "desc"}}], | ||
"query": { | ||
"bool": { | ||
"must": [ | ||
{ | ||
"query_string": { | ||
"query": "account_history.account: )" + account_id_string + range + R"(" | ||
} | ||
}, | ||
{ | ||
"range": { | ||
"block_data.block_time": { | ||
"gte": "now-20y", | ||
"lte": "now" | ||
} | ||
} | ||
} | ||
] | ||
} | ||
} | ||
} | ||
)"; | ||
|
||
auto es = prepareHistoryQuery(query); | ||
|
||
vector<operation_history_object> result; | ||
|
||
if(!graphene::utilities::checkES(es)) | ||
return result; | ||
|
||
const auto response = graphene::utilities::simpleQuery(es); | ||
variant variant_response = fc::json::from_string(response); | ||
|
||
const auto hits = variant_response["hits"]["total"]; | ||
const auto size = std::min(static_cast<uint32_t>(hits.as_uint64()), limit); | ||
|
||
for(unsigned i=0; i<size; i++) | ||
{ | ||
const auto source = variant_response["hits"]["hits"][size_t(i)]["_source"]; | ||
result.push_back(fromEStoOperation(source)); | ||
} | ||
return result; | ||
} | ||
|
||
operation_history_object elasticsearch_plugin::fromEStoOperation(variant source) | ||
{ | ||
operation_history_object result; | ||
|
||
const auto operation_id = source["account_history"]["operation_id"]; | ||
fc::from_variant( operation_id, result.id, GRAPHENE_MAX_NESTED_OBJECTS ); | ||
|
||
const auto op = fc::json::from_string(source["operation_history"]["op"].as_string()); | ||
fc::from_variant( op, result.op, GRAPHENE_MAX_NESTED_OBJECTS ); | ||
|
||
const auto operation_result = fc::json::from_string(source["operation_history"]["operation_result"].as_string()); | ||
fc::from_variant( operation_result, result.result, GRAPHENE_MAX_NESTED_OBJECTS ); | ||
|
||
result.block_num = source["block_data"]["block_num"].as_uint64(); | ||
result.trx_in_block = source["operation_history"]["trx_in_block"].as_uint64(); | ||
result.op_in_trx = source["operation_history"]["op_in_trx"].as_uint64(); | ||
result.trx_in_block = source["operation_history"]["virtual_op"].as_uint64(); | ||
|
||
return result; | ||
} | ||
|
||
graphene::utilities::ES elasticsearch_plugin::prepareHistoryQuery(string query) | ||
{ | ||
CURL *curl; | ||
curl = curl_easy_init(); | ||
|
||
graphene::utilities::ES es; | ||
es.curl = curl; | ||
es.elasticsearch_url = my->_elasticsearch_node_url; | ||
es.index_prefix = my->_elasticsearch_index_prefix; | ||
es.endpoint = es.index_prefix + "*/data/_search"; | ||
es.query = query; | ||
|
||
return es; | ||
} | ||
|
||
} } |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please capture only what's necessary.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Also, don't use this if mode is
only_save
.There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Does this actually work as intended? That would be cool.
We should load-test this before putting it into production.