CMS 3D CMS Logo

List of all members | Classes | Public Member Functions | Static Public Member Functions | Private Member Functions | Private Attributes
TritonService Class Reference

#include <TritonService.h>

Classes

struct  FallbackOpts
 
struct  Model
 
struct  Module
 
struct  Server
 

Public Member Functions

void addModel (const std::string &modelName, const std::string &path)
 
void notifyCallStatus (bool status) const
 
const std::string & pid () const
 
Server serverInfo (const std::string &model, const std::string &preferred="") const
 
 TritonService (const edm::ParameterSet &pset, edm::ActivityRegistry &areg)
 
 ~TritonService ()=default
 

Static Public Member Functions

static void fillDescriptions (edm::ConfigurationDescriptions &descriptions)
 

Private Member Functions

void postEndJob ()
 
void postModuleConstruction (edm::ModuleDescription const &)
 
void preallocate (edm::service::SystemBounds const &)
 
void preBeginJob (edm::PathsAndConsumesOfModulesBase const &, edm::ProcessContext const &)
 
void preModuleConstruction (edm::ModuleDescription const &)
 
void preModuleDestruction (edm::ModuleDescription const &)
 
template<typename LOG >
void printFallbackServerLog () const
 

Private Attributes

bool allowAddModel_
 
std::atomic< int > callFails_
 
unsigned currentModuleId_
 
FallbackOpts fallbackOpts_
 
std::unordered_map< std::string, Modelmodels_
 
std::unordered_map< unsigned, Modulemodules_
 
int numberOfThreads_
 
std::string pid_
 
std::unordered_map< std::string, Serverservers_
 
bool startedFallback_
 
std::unordered_map< std::string, ModelunservedModels_
 
bool verbose_
 

Detailed Description

Definition at line 31 of file TritonService.h.

Constructor & Destructor Documentation

◆ TritonService()

TritonService::TritonService ( const edm::ParameterSet pset,
edm::ActivityRegistry areg 
)

Definition at line 58 of file TritonService.cc.

References HLTObjectMonitor_Client_cff::client, TritonService::FallbackOpts::enable, Exception, TritonService::Server::fallbackAddress, TritonService::Server::fallbackName, fallbackOpts_, LocalCPU, LocalGPU, ecalBarrelClusterFastTimer_cfi::modelName, models_, mps_check::msg, postEndJob(), postModuleConstruction(), preallocate(), preBeginJob(), preModuleConstruction(), preModuleDestruction(), muonDTDigis_cfi::pset, Remote, contentValuesFiles::server, servers_, AlCaHLTBitMon_QueryRunRegistry::string, TRITON_THROW_IF_ERROR, tier0::unique(), TritonService::FallbackOpts::useGPU, verbose_, edm::ActivityRegistry::watchPostEndJob(), edm::ActivityRegistry::watchPostModuleConstruction(), edm::ActivityRegistry::watchPreallocate(), edm::ActivityRegistry::watchPreBeginJob(), edm::ActivityRegistry::watchPreModuleConstruction(), and edm::ActivityRegistry::watchPreModuleDestruction().

59  : verbose_(pset.getUntrackedParameter<bool>("verbose")),
60  fallbackOpts_(pset.getParameterSet("fallback")),
62  allowAddModel_(false),
63  startedFallback_(false),
64  callFails_(0),
65  pid_(std::to_string(::getpid())) {
66  //module construction is assumed to be serial (correct at the time this code was written)
67 
69 
73  //fallback server will be launched (if needed) before beginJob
76 
77  //include fallback server in set if enabled
78  if (fallbackOpts_.enable) {
79  auto serverType = TritonServerType::Remote;
80  if (!fallbackOpts_.useGPU)
81  serverType = TritonServerType::LocalCPU;
82 #ifdef TRITON_ENABLE_GPU
83  else
84  serverType = TritonServerType::LocalGPU;
85 #endif
86 
87  servers_.emplace(std::piecewise_construct,
88  std::forward_as_tuple(Server::fallbackName),
89  std::forward_as_tuple(Server::fallbackName, Server::fallbackAddress, serverType));
90  }
91 
92  //loop over input servers: check which models they have
94  if (verbose_)
95  msg = "List of models for each server:\n";
96  for (const auto& serverPset : pset.getUntrackedParameterSetVector("servers")) {
97  const std::string& serverName(serverPset.getUntrackedParameter<std::string>("name"));
98  //ensure uniqueness
99  auto [sit, unique] = servers_.emplace(serverName, serverPset);
100  if (!unique)
101  throw cms::Exception("DuplicateServer")
102  << "TritonService: Not allowed to specify more than one server with same name (" << serverName << ")";
103  auto& server(sit->second);
104 
105  std::unique_ptr<tc::InferenceServerGrpcClient> client;
107  tc::InferenceServerGrpcClient::Create(&client, server.url, false, server.useSsl, server.sslOptions),
108  "TritonService(): unable to create inference context for " + serverName + " (" + server.url + ")",
109  false);
110 
111  if (verbose_) {
112  inference::ServerMetadataResponse serverMetaResponse;
113  TRITON_THROW_IF_ERROR(client->ServerMetadata(&serverMetaResponse),
114  "TritonService(): unable to get metadata for " + serverName + " (" + server.url + ")",
115  false);
116  edm::LogInfo("TritonService") << "Server " << serverName << ": url = " << server.url
117  << ", version = " << serverMetaResponse.version();
118  }
119 
120  inference::RepositoryIndexResponse repoIndexResponse;
121  TRITON_THROW_IF_ERROR(client->ModelRepositoryIndex(&repoIndexResponse),
122  "TritonService(): unable to get repository index for " + serverName + " (" + server.url + ")",
123  false);
124 
125  //servers keep track of models and vice versa
126  if (verbose_)
127  msg += serverName + ": ";
128  for (const auto& modelIndex : repoIndexResponse.models()) {
129  const auto& modelName = modelIndex.name();
130  auto mit = models_.find(modelName);
131  if (mit == models_.end())
132  mit = models_.emplace(modelName, "").first;
133  auto& modelInfo(mit->second);
134  modelInfo.servers.insert(serverName);
135  server.models.insert(modelName);
136  if (verbose_)
137  msg += modelName + ", ";
138  }
139  if (verbose_)
140  msg += "\n";
141  }
142  if (verbose_)
143  edm::LogInfo("TritonService") << msg;
144 }
void watchPostModuleConstruction(PostModuleConstruction::slot_type const &iSlot)
std::unordered_map< std::string, Model > models_
void watchPreallocate(Preallocate::slot_type const &iSlot)
void watchPostEndJob(PostEndJob::slot_type const &iSlot)
void watchPreModuleConstruction(PreModuleConstruction::slot_type const &iSlot)
void postModuleConstruction(edm::ModuleDescription const &)
void watchPreModuleDestruction(PreModuleDestruction::slot_type const &iSlot)
unsigned currentModuleId_
void preallocate(edm::service::SystemBounds const &)
static const std::string fallbackAddress
Definition: TritonService.h:91
static std::string to_string(const XMLCh *ch)
T getUntrackedParameter(std::string const &, T const &) const
FallbackOpts fallbackOpts_
def unique(seq, keepstr=True)
Definition: tier0.py:24
std::string pid_
void preBeginJob(edm::PathsAndConsumesOfModulesBase const &, edm::ProcessContext const &)
#define TRITON_THROW_IF_ERROR(X, MSG, NOTIFY)
Definition: triton_utils.h:78
void preModuleConstruction(edm::ModuleDescription const &)
Log< level::Info, false > LogInfo
void preModuleDestruction(edm::ModuleDescription const &)
std::atomic< int > callFails_
tuple msg
Definition: mps_check.py:286
void watchPreBeginJob(PreBeginJob::slot_type const &iSlot)
convenience function for attaching to signal
std::unordered_map< std::string, Server > servers_
static const std::string fallbackName
Definition: TritonService.h:90
Definition: server.py:1

◆ ~TritonService()

TritonService::~TritonService ( )
default

Member Function Documentation

◆ addModel()

void TritonService::addModel ( const std::string &  modelName,
const std::string &  path 
)

Definition at line 155 of file TritonService.cc.

References allowAddModel_, currentModuleId_, Exception, ecalBarrelClusterFastTimer_cfi::modelName, models_, modules_, castor_dqm_sourceclient_file_cfg::path, and unservedModels_.

Referenced by TritonEDFilterT< G, Capabilities >::initializeGlobalCache(), TritonEDProducerT< G, Capabilities >::initializeGlobalCache(), and TritonOneEDAnalyzer< Capabilities >::TritonOneEDAnalyzer().

155  {
156  //should only be called in module constructors
157  if (!allowAddModel_)
158  throw cms::Exception("DisallowedAddModel")
159  << "TritonService: Attempt to call addModel() outside of module constructors";
160  //if model is not in the list, then no specified server provides it
161  auto mit = models_.find(modelName);
162  if (mit == models_.end()) {
163  auto& modelInfo(unservedModels_.emplace(modelName, path).first->second);
164  modelInfo.modules.insert(currentModuleId_);
165  //only keep track of modules that need unserved models
167  }
168 }
std::unordered_map< std::string, Model > models_
std::unordered_map< std::string, Model > unservedModels_
unsigned currentModuleId_
std::unordered_map< unsigned, Module > modules_

◆ fillDescriptions()

void TritonService::fillDescriptions ( edm::ConfigurationDescriptions descriptions)
static

Definition at line 362 of file TritonService.cc.

References edm::ParameterSetDescription::addUntracked(), edm::ConfigurationDescriptions::addWithDefaultLabel(), submitPVResolutionJobs::desc, and AlCaHLTBitMon_QueryRunRegistry::string.

362  {
364  desc.addUntracked<bool>("verbose", false);
365 
367  validator.addUntracked<std::string>("name");
368  validator.addUntracked<std::string>("address");
369  validator.addUntracked<unsigned>("port");
370  validator.addUntracked<bool>("useSsl", false);
371  validator.addUntracked<std::string>("rootCertificates", "");
372  validator.addUntracked<std::string>("privateKey", "");
373  validator.addUntracked<std::string>("certificateChain", "");
374 
375  desc.addVPSetUntracked("servers", validator, {});
376 
377  edm::ParameterSetDescription fallbackDesc;
378  fallbackDesc.addUntracked<bool>("enable", false);
379  fallbackDesc.addUntracked<bool>("debug", false);
380  fallbackDesc.addUntracked<bool>("verbose", false);
381  fallbackDesc.addUntracked<bool>("useDocker", false);
382  fallbackDesc.addUntracked<bool>("useGPU", false);
383  fallbackDesc.addUntracked<int>("retries", -1);
384  fallbackDesc.addUntracked<int>("wait", -1);
385  fallbackDesc.addUntracked<std::string>("instanceBaseName", "triton_server_instance");
386  fallbackDesc.addUntracked<std::string>("instanceName", "");
387  fallbackDesc.addUntracked<std::string>("tempDir", "");
388  fallbackDesc.addUntracked<std::string>("imageName", "");
389  fallbackDesc.addUntracked<std::string>("sandboxName", "");
390  desc.add<edm::ParameterSetDescription>("fallback", fallbackDesc);
391 
392  descriptions.addWithDefaultLabel(desc);
393 }
void addWithDefaultLabel(ParameterSetDescription const &psetDescription)
ParameterDescriptionBase * addUntracked(U const &iLabel, T const &value)

◆ notifyCallStatus()

void TritonService::notifyCallStatus ( bool  status) const

Definition at line 303 of file TritonService.cc.

References callFails_, and mps_update::status.

Referenced by TritonClient::evaluate(), and TritonException::TritonException().

303  {
304  if (status)
305  --callFails_;
306  else
307  ++callFails_;
308 }
std::atomic< int > callFails_

◆ pid()

const std::string& TritonService::pid ( ) const
inline

Definition at line 115 of file TritonService.h.

References pid_.

Referenced by TritonClient::TritonClient().

115 { return pid_; }
std::string pid_

◆ postEndJob()

void TritonService::postEndJob ( )
private

Definition at line 310 of file TritonService.cc.

References callFails_, mps_check::command, TritonService::FallbackOpts::command, Exception, fallbackOpts_, dqm-mbProfile::format, or, convertSQLitetoXML_cfg::output, startedFallback_, AlCaHLTBitMon_QueryRunRegistry::string, and verbose_.

Referenced by TritonService().

310  {
311  if (!startedFallback_)
312  return;
313 
315  //prevent log cleanup during server stop
316  if (callFails_ > 0)
317  command += " -c";
318  command += " stop";
319  if (verbose_)
320  edm::LogInfo("TritonService") << command;
321 
322  const auto& [output, rv] = execSys(command);
323  if (rv != 0 or callFails_ > 0) {
324  //print logs if cmsRun is currently exiting because of a TritonException
325  edm::LogError("TritonService") << output;
326  printFallbackServerLog<edm::LogError>();
327  if (rv != 0) {
328  std::string stopCat("FallbackFailed");
329  std::string stopMsg = fmt::format("TritonService: Stopping the fallback server failed with exit code {}", rv);
330  //avoid throwing if the stack is already unwinding
331  if (callFails_ > 0)
332  edm::LogWarning(stopCat) << stopMsg;
333  else
334  throw cms::Exception(stopCat) << stopMsg;
335  }
336  } else if (verbose_) {
337  edm::LogInfo("TritonService") << output;
338  printFallbackServerLog<edm::LogInfo>();
339  }
340 }
Log< level::Error, false > LogError
FallbackOpts fallbackOpts_
The Signals That Services Can Subscribe To This is based on ActivityRegistry and is current per Services can connect to the signals distributed by the ActivityRegistry in order to monitor the activity of the application Each possible callback has some defined which we here list in angle e< void, edm::EventID const &, edm::Timestamp const & > We also list in braces which AR_WATCH_USING_METHOD_ is used for those or
Definition: Activities.doc:12
Log< level::Info, false > LogInfo
std::atomic< int > callFails_
list command
Definition: mps_check.py:25
Log< level::Warning, false > LogWarning

◆ postModuleConstruction()

void TritonService::postModuleConstruction ( edm::ModuleDescription const &  desc)
private

Definition at line 170 of file TritonService.cc.

References allowAddModel_.

Referenced by TritonService().

170 { allowAddModel_ = false; }

◆ preallocate()

void TritonService::preallocate ( edm::service::SystemBounds const &  bounds)
private

Definition at line 146 of file TritonService.cc.

References numberOfThreads_.

Referenced by TritonService().

146  {
147  numberOfThreads_ = bounds.maxNumberOfThreads();
148 }

◆ preBeginJob()

void TritonService::preBeginJob ( edm::PathsAndConsumesOfModulesBase const &  ,
edm::ProcessContext const &   
)
private

Definition at line 215 of file TritonService.cc.

References mps_check::command, TritonService::FallbackOpts::command, edm::createGlobalIdentifier(), TritonService::FallbackOpts::debug, TritonService::FallbackOpts::enable, Exception, TritonService::Server::fallbackName, fallbackOpts_, ALPAKA_ACCELERATOR_NAMESPACE::caPixelDoublets::if(), TritonService::FallbackOpts::imageName, TritonService::FallbackOpts::instanceName, ReggeGribovPartonMC_EposLHC_2760GeV_PbPb_cfi::model, ecalBarrelClusterFastTimer_cfi::modelName, models_, mps_check::msg, numberOfThreads_, or, convertSQLitetoXML_cfg::output, pid_, TritonService::FallbackOpts::retries, TritonService::FallbackOpts::sandboxName, contentValuesFiles::server, servers_, startedFallback_, AlCaHLTBitMon_QueryRunRegistry::string, TritonService::FallbackOpts::tempDir, to_string(), unservedModels_, TritonService::FallbackOpts::useDocker, TritonService::FallbackOpts::useGPU, TritonService::FallbackOpts::verbose, verbose_, and TritonService::FallbackOpts::wait.

Referenced by TritonService().

215  {
216  //only need fallback if there are unserved models
217  if (!fallbackOpts_.enable or unservedModels_.empty())
218  return;
219 
221  if (verbose_)
222  msg = "List of models for fallback server: ";
223  //all unserved models are provided by fallback server
224  auto& server(servers_.find(Server::fallbackName)->second);
225  for (const auto& [modelName, model] : unservedModels_) {
226  auto& modelInfo(models_.emplace(modelName, model).first->second);
227  modelInfo.servers.insert(Server::fallbackName);
228  server.models.insert(modelName);
229  if (verbose_)
230  msg += modelName + ", ";
231  }
232  if (verbose_)
233  edm::LogInfo("TritonService") << msg;
234 
235  //assemble server start command
236  fallbackOpts_.command = "cmsTriton -P -1 -p " + pid_;
237  if (fallbackOpts_.debug)
238  fallbackOpts_.command += " -c";
240  fallbackOpts_.command += " -v";
242  fallbackOpts_.command += " -d";
243  if (fallbackOpts_.useGPU)
244  fallbackOpts_.command += " -g";
245  if (!fallbackOpts_.instanceName.empty())
247  if (fallbackOpts_.retries >= 0)
249  if (fallbackOpts_.wait >= 0)
251  for (const auto& [modelName, model] : unservedModels_) {
252  fallbackOpts_.command += " -m " + model.path;
253  }
254  std::string thread_string = " -I " + std::to_string(numberOfThreads_);
255  fallbackOpts_.command += thread_string;
256  if (!fallbackOpts_.imageName.empty())
258  if (!fallbackOpts_.sandboxName.empty())
260  //don't need this anymore
261  unservedModels_.clear();
262 
263  //get a random temporary directory if none specified
264  if (fallbackOpts_.tempDir.empty()) {
265  auto tmp_dir_path{std::filesystem::temp_directory_path() /= edm::createGlobalIdentifier()};
266  fallbackOpts_.tempDir = tmp_dir_path.string();
267  }
268  //special case ".": use script default (temp dir = .$instanceName)
269  if (fallbackOpts_.tempDir != ".")
271 
273 
274  if (fallbackOpts_.debug)
275  edm::LogInfo("TritonService") << "Fallback server temporary directory: " << fallbackOpts_.tempDir;
276  if (verbose_)
277  edm::LogInfo("TritonService") << command;
278 
279  //mark as started before executing in case of ctrl+c while command is running
280  startedFallback_ = true;
281  const auto& [output, rv] = execSys(command);
282  if (rv != 0) {
283  edm::LogError("TritonService") << output;
284  printFallbackServerLog<edm::LogError>();
285  throw cms::Exception("FallbackFailed")
286  << "TritonService: Starting the fallback server failed with exit code " << rv;
287  } else if (verbose_)
288  edm::LogInfo("TritonService") << output;
289  //get the port
290  const std::string& portIndicator("CMS_TRITON_GRPC_PORT: ");
291  //find last instance in log in case multiple ports were tried
292  auto pos = output.rfind(portIndicator);
293  if (pos != std::string::npos) {
294  auto pos2 = pos + portIndicator.size();
295  auto pos3 = output.find('\n', pos2);
296  const auto& portNum = output.substr(pos2, pos3 - pos2);
297  server.url += ":" + portNum;
298  } else
299  throw cms::Exception("FallbackFailed") << "TritonService: Unknown port for fallback server, log follows:\n"
300  << output;
301 }
std::unordered_map< std::string, Model > models_
std::unordered_map< std::string, Model > unservedModels_
Log< level::Error, false > LogError
static std::string to_string(const XMLCh *ch)
FallbackOpts fallbackOpts_
The Signals That Services Can Subscribe To This is based on ActivityRegistry and is current per Services can connect to the signals distributed by the ActivityRegistry in order to monitor the activity of the application Each possible callback has some defined which we here list in angle e< void, edm::EventID const &, edm::Timestamp const & > We also list in braces which AR_WATCH_USING_METHOD_ is used for those or
Definition: Activities.doc:12
std::string pid_
std::string createGlobalIdentifier(bool binary=false)
Log< level::Info, false > LogInfo
tuple msg
Definition: mps_check.py:286
list command
Definition: mps_check.py:25
Definition: output.py:1
if(threadIdxLocalY==0 &&threadIdxLocalX==0)
std::unordered_map< std::string, Server > servers_
static const std::string fallbackName
Definition: TritonService.h:90
Definition: server.py:1

◆ preModuleConstruction()

void TritonService::preModuleConstruction ( edm::ModuleDescription const &  desc)
private

Definition at line 150 of file TritonService.cc.

References allowAddModel_, currentModuleId_, and submitPVResolutionJobs::desc.

Referenced by TritonService().

150  {
151  currentModuleId_ = desc.id();
152  allowAddModel_ = true;
153 }
unsigned currentModuleId_

◆ preModuleDestruction()

void TritonService::preModuleDestruction ( edm::ModuleDescription const &  desc)
private

Definition at line 172 of file TritonService.cc.

References submitPVResolutionJobs::desc, modules_, and unservedModels_.

Referenced by TritonService().

172  {
173  //remove destructed modules from unserved list
174  if (unservedModels_.empty())
175  return;
176  auto id = desc.id();
177  auto oit = modules_.find(id);
178  if (oit != modules_.end()) {
179  const auto& moduleInfo(oit->second);
180  auto mit = unservedModels_.find(moduleInfo.model);
181  if (mit != unservedModels_.end()) {
182  auto& modelInfo(mit->second);
183  modelInfo.modules.erase(id);
184  //remove a model if it is no longer needed by any modules
185  if (modelInfo.modules.empty())
186  unservedModels_.erase(mit);
187  }
188  modules_.erase(oit);
189  }
190 }
std::unordered_map< std::string, Model > unservedModels_
std::unordered_map< unsigned, Module > modules_

◆ printFallbackServerLog()

template<typename LOG >
void TritonService::printFallbackServerLog ( ) const
private

Definition at line 343 of file TritonService.cc.

References fallbackOpts_, timingPdfMaker::infile, TritonService::FallbackOpts::instanceName, LOG, heppy_check::logName, and TritonService::FallbackOpts::tempDir.

343  {
344  std::vector<std::string> logNames{"log_" + fallbackOpts_.instanceName + ".log"};
345  //cmsTriton script moves log from temp to current dir in verbose mode or in some cases when auto_stop is called
346  // -> check both places
347  logNames.push_back(fallbackOpts_.tempDir + "/" + logNames[0]);
348  bool foundLog = false;
349  for (const auto& logName : logNames) {
350  std::ifstream infile(logName);
351  if (infile.is_open()) {
352  LOG("TritonService") << "TritonService: server log " << logName << "\n" << infile.rdbuf();
353  foundLog = true;
354  break;
355  }
356  }
357  if (!foundLog)
358  LOG("TritonService") << "TritonService: could not find server log " << logNames[0] << " in current directory or "
360 }
#define LOG(A)
FallbackOpts fallbackOpts_

◆ serverInfo()

TritonService::Server TritonService::serverInfo ( const std::string &  model,
const std::string &  preferred = "" 
) const

Definition at line 193 of file TritonService.cc.

References ReggeGribovPartonMC_EposLHC_2760GeV_PbPb_cfi::model, models_, contentValuesFiles::server, and servers_.

Referenced by TritonClient::TritonClient().

193  {
194  auto mit = models_.find(model);
195  if (mit == models_.end())
196  throw cms::Exception("MissingModel") << "TritonService: There are no servers that provide model " << model;
197  const auto& modelInfo(mit->second);
198  const auto& modelServers = modelInfo.servers;
199 
200  auto msit = modelServers.end();
201  if (!preferred.empty()) {
202  msit = modelServers.find(preferred);
203  //todo: add a "strict" parameter to stop execution if preferred server isn't found?
204  if (msit == modelServers.end())
205  edm::LogWarning("PreferredServer") << "Preferred server " << preferred << " for model " << model
206  << " not available, will choose another server";
207  }
208  const auto& serverName(msit == modelServers.end() ? *modelServers.begin() : preferred);
209 
210  //todo: use some algorithm to select server rather than just picking arbitrarily
211  const auto& server(servers_.find(serverName)->second);
212  return server;
213 }
std::unordered_map< std::string, Model > models_
std::unordered_map< std::string, Server > servers_

Member Data Documentation

◆ allowAddModel_

bool TritonService::allowAddModel_
private

Definition at line 135 of file TritonService.h.

Referenced by addModel(), postModuleConstruction(), and preModuleConstruction().

◆ callFails_

std::atomic<int> TritonService::callFails_
mutableprivate

Definition at line 137 of file TritonService.h.

Referenced by notifyCallStatus(), and postEndJob().

◆ currentModuleId_

unsigned TritonService::currentModuleId_
private

Definition at line 134 of file TritonService.h.

Referenced by addModel(), and preModuleConstruction().

◆ fallbackOpts_

FallbackOpts TritonService::fallbackOpts_
private

Definition at line 133 of file TritonService.h.

Referenced by postEndJob(), preBeginJob(), printFallbackServerLog(), and TritonService().

◆ models_

std::unordered_map<std::string, Model> TritonService::models_
private

Definition at line 142 of file TritonService.h.

Referenced by addModel(), preBeginJob(), serverInfo(), and TritonService().

◆ modules_

std::unordered_map<unsigned, Module> TritonService::modules_
private

Definition at line 143 of file TritonService.h.

Referenced by addModel(), and preModuleDestruction().

◆ numberOfThreads_

int TritonService::numberOfThreads_
private

Definition at line 144 of file TritonService.h.

Referenced by preallocate(), and preBeginJob().

◆ pid_

std::string TritonService::pid_
private

Definition at line 138 of file TritonService.h.

Referenced by pid(), and preBeginJob().

◆ servers_

std::unordered_map<std::string, Server> TritonService::servers_
private

Definition at line 141 of file TritonService.h.

Referenced by preBeginJob(), serverInfo(), and TritonService().

◆ startedFallback_

bool TritonService::startedFallback_
private

Definition at line 136 of file TritonService.h.

Referenced by postEndJob(), and preBeginJob().

◆ unservedModels_

std::unordered_map<std::string, Model> TritonService::unservedModels_
private

Definition at line 139 of file TritonService.h.

Referenced by addModel(), preBeginJob(), and preModuleDestruction().

◆ verbose_

bool TritonService::verbose_
private

Definition at line 132 of file TritonService.h.

Referenced by postEndJob(), preBeginJob(), and TritonService().