WorkflowTask.h
1 
10 #ifndef WRENCH_WORKFLOWTASK_H
11 #define WRENCH_WORKFLOWTASK_H
12 
13 #include <map>
14 #include <stack>
15 #include <set>
16 
17 #include "wrench/workflow/job/WorkflowJob.h"
18 #include "wrench/workflow/WorkflowFile.h"
19 
20 #include <boost/graph/adjacency_list.hpp>
21 
22 namespace wrench {
23 
27  class WorkflowTask {
28 
29  public:
30  std::string getID() const;
31 
32  double getFlops() const;
33 
34  unsigned long getMinNumCores() const;
35 
36  unsigned long getMaxNumCores() const;
37 
38  double getParallelEfficiency() const;
39 
40  double getMemoryRequirement() const;
41 
42  unsigned long getNumberOfChildren() const;
43 
44  std::vector<WorkflowTask *> getChildren() const;
45 
46  unsigned long getNumberOfParents() const;
47 
48  std::vector<WorkflowTask *> getParents() const;
49 
50  void addInputFile(WorkflowFile *file);
51 
52  void addOutputFile(WorkflowFile *file);
53 
54  unsigned int getFailureCount();
55 
56 
57  /***********************/
59  /***********************/
60 
62  enum State {
73  };
74 
75  static std::string stateToString(WorkflowTask::State state);
76 
77  WorkflowJob *getJob() const;
78 
79  Workflow *getWorkflow() const;
80 
81  std::string getClusterID() const;
82 
83  void setClusterID(std::string);
84 
85  void setPriority(long);
86 
87  unsigned long getPriority() const;
88 
89  void setAverageCPU(double);
90 
91  double getAverageCPU() const;
92 
93  void setBytesRead(unsigned long);
94 
95  unsigned long getBytesRead() const;
96 
97  void setBytesWritten(unsigned long);
98 
99  unsigned long getBytesWritten() const;
100 
101  std::vector<WorkflowFile *> getInputFiles();
102 
103  std::vector<WorkflowFile *> getOutputFiles();
104 
105  unsigned long getTopLevel();
106 
107  double getStartDate();
108 
109  double getEndDate();
110 
111  double getFailureDate();
112 
113  double getTerminationDate();
114 
115  double getReadInputStartDate();
116 
117  double getReadInputEndDate();
118 
119  double getComputationStartDate();
120 
121  double getComputationEndDate();
122 
123  double getWriteOutputStartDate();
124 
125  double getWriteOutputEndDate();
126 
127  unsigned long getNumCoresAllocated();
128 
129  struct WorkflowTaskExecution;
130 
131  std::stack<WorkflowTaskExecution> getExecutionHistory();
132 
133  std::string getExecutionHost();
134 
136 
137  std::string getColor();
138 
139  void setColor(std::string);
140 
141  /***********************/
143  /***********************/
144 
145 
146  /***********************/
148  /***********************/
149 
152  TASK_NOT_READY,
153  TASK_READY,
154  TASK_RUNNING,
155  TASK_COMPLETED,
156  TASK_FAILED
157  };
158 
159  static std::string stateToString(WorkflowTask::InternalState state);
160 
161  unsigned long updateTopLevel();
162 
164 
166 
168 
170 
172 
173  void setJob(WorkflowJob *job);
174 
175  void setStartDate(double date);
176 
177  void setEndDate(double date);
178 
179  void setReadInputStartDate(double date);
180 
181  void setReadInputEndDate(double date);
182 
183  void setComputationStartDate(double date);
184 
185  void setComputationEndDate(double date);
186 
187  void setWriteOutputStartDate(double date);
188 
189  void setWriteOutputEndDate(double date);
190 
191  void setFailureDate(double date);
192 
193  void setTerminationDate(double date);
194 
195  void incrementFailureCount();
196 
197  void setExecutionHost(std::string hostname);
198 
199  void setNumCoresAllocated(unsigned long num_cores);
200 
206  double task_start = -1.0;
208  double read_input_start = -1.0;
210  double read_input_end = -1.0;
212  double computation_start = -1.0;
214  double computation_end = -1.0;
216  double write_output_start = -1.0;
218  double write_output_end = -1.0;
220  double task_end = -1.0;
222  double task_failed = -1.0;
224  double task_terminated = -1.0;
225 
227  std::string execution_host = "";
229  unsigned long num_cores_allocated = 0;
230 
237 
238  };
239 
240  /***********************/
242  /***********************/
243 
244  private:
245  friend class Workflow;
246 
247  std::string id; // Task ID
248  std::string cluster_id; // ID for clustered task
249  std::string color; // A RGB color formatted as "#rrggbb"
250  double flops; // Number of flops
251  double average_cpu = -1; // Average CPU utilization
252  unsigned long bytes_read = -1; // Total bytes read in KB
253  unsigned long bytes_written = -1; // Total bytes written in KB
254  unsigned long min_num_cores;
255  unsigned long max_num_cores;
256  double parallel_efficiency;
257  double memory_requirement;
258  unsigned long priority = 0; // Task priority
259  unsigned long toplevel; // 0 if entry task
260  unsigned int failure_count = 0; // Number of times the tasks has failed
261  std::string execution_host; // Host on which the task executed ("" if not executed successfully - yet)
262  State visible_state; // To be exposed to developer level
263  State upcoming_visible_state; // A visible state that will become active once a WMS has process a previously sent workflow execution event
264  InternalState internal_state; // Not to be exposed to developer level
265 
266  Workflow *workflow; // Containing workflow
267 
268  std::map<std::string, WorkflowFile *> output_files; // List of output files
269  std::map<std::string, WorkflowFile *> input_files; // List of input files
270 
271  // Private constructor (called by Workflow)
272  WorkflowTask(std::string id,
273  double t,
274  unsigned long min_num_cores,
275  unsigned long max_num_cores,
276  double parallel_efficiency,
277  double memory_requirement);
278 
279  // Containing job
280  WorkflowJob *job;
281 
282  std::stack<WorkflowTaskExecution> execution_history;
283 
284  friend class DagOfTasks;
285  };
286 };
287 
288 #endif //WRENCH_WORKFLOWTASK_H
@ PENDING
Pending (has been submitted to a compute service)
Definition: WorkflowTask.h:68
void setUpcomingState(WorkflowTask::State)
Set the upcoming visible state of the task.
Definition: WorkflowTask.cpp:346
@ READY
Ready (parents have completed)
Definition: WorkflowTask.h:66
double getComputationEndDate()
Get the task's most recent computation end date.
Definition: WorkflowTask.cpp:668
An internal class that uses the Boost Graph Library to implement a DAG of WorkflowTask objects.
Definition: DagOfTasks.h:40
std::string execution_host
Task's execution host.
Definition: WorkflowTask.h:227
void incrementFailureCount()
Increment the failure count of a task.
Definition: WorkflowTask.cpp:610
double write_output_end
Task's write output end time.
Definition: WorkflowTask.h:218
@ UNKNOWN
Some Unknown state (should not happen)
Definition: WorkflowTask.h:72
WorkflowTask::State getState() const
Get the state of the task.
Definition: WorkflowTask.cpp:205
std::string getID() const
Get the id of the task.
Definition: WorkflowTask.cpp:115
State
Task states.
Definition: WorkflowTask.h:62
unsigned long updateTopLevel()
Update the task's top level (looking only at the parents, and updating children)
Definition: WorkflowTask.cpp:724
void setComputationEndDate(double date)
Set the date when the computation portion of a WorkflowTask has ended.
Definition: WorkflowTask.cpp:492
double getAverageCPU() const
Get the task average CPU usage.
Definition: WorkflowTask.cpp:404
void setClusterID(std::string)
Set the cluster id for the task.
Definition: WorkflowTask.cpp:380
void setExecutionHost(std::string hostname)
Sets the host on which this task is running.
Definition: WorkflowTask.cpp:772
double read_input_end
Task's read input end time.
Definition: WorkflowTask.h:210
std::vector< WorkflowTask * > getChildren() const
Get the children of a task.
Definition: WorkflowTask.cpp:178
WorkflowTask::InternalState getInternalState() const
Get the state of the task (as known to the "internal" layer)
Definition: WorkflowTask.cpp:223
void setTerminationDate(double date)
Set the date when the task was terminated.
Definition: WorkflowTask.cpp:580
std::vector< WorkflowFile * > getOutputFiles()
Get the list of output WorkflowFile objects for the task.
Definition: WorkflowTask.cpp:631
void addInputFile(WorkflowFile *file)
Add an input file to the task.
Definition: WorkflowTask.cpp:55
void setNumCoresAllocated(unsigned long num_cores)
Sets the number of cores allocated for this task.
Definition: WorkflowTask.cpp:785
double write_output_start
Task's write output start time.
Definition: WorkflowTask.h:216
void setComputationStartDate(double date)
Set the date when the computation portion of a WorkflowTask has begun.
Definition: WorkflowTask.cpp:477
unsigned long getMaxNumCores() const
Get the maximum number of cores that the task can use.
Definition: WorkflowTask.cpp:142
double task_end
Task's end time.
Definition: WorkflowTask.h:220
unsigned long getNumberOfChildren() const
Get the number of children of a task.
Definition: WorkflowTask.cpp:169
std::string getColor()
Get the task's color ("" if none)
Definition: WorkflowTask.cpp:798
double read_input_start
Task's read input start time.
Definition: WorkflowTask.h:208
void setBytesWritten(unsigned long)
Set the number of bytes written by the task.
Definition: WorkflowTask.cpp:444
void setReadInputEndDate(double date)
Set the date when the read input portion of a WorkflowTask has completed.
Definition: WorkflowTask.cpp:522
double getParallelEfficiency() const
Get the parallel efficiency of the task.
Definition: WorkflowTask.cpp:151
void setFailureDate(double date)
Set the date when the task has failed.
Definition: WorkflowTask.cpp:566
double getWriteOutputStartDate()
Get the task's most recent write output start date.
Definition: WorkflowTask.cpp:692
double getStartDate()
Get the task's most recent start date.
Definition: WorkflowTask.cpp:644
Definition: Alarm.cpp:20
A workflow (to be executed by a WMS)
Definition: Workflow.h:33
@ COMPLETED
Completed (successfully completed)
Definition: WorkflowTask.h:70
static std::string stateToString(WorkflowTask::State state)
Convert task state to a string (useful for output, debugging, logging, etc.)
Definition: WorkflowTask.cpp:232
double getTerminationDate()
Get the tasks's most recent termination date (when it was explicitly requested to be terminated by th...
Definition: WorkflowTask.cpp:716
void setPriority(long)
Set the task priority.
Definition: WorkflowTask.cpp:396
unsigned long num_cores_allocated
Task's number of allocated cores.
Definition: WorkflowTask.h:229
unsigned long getMinNumCores() const
Get the minimum number of cores required for running the task.
Definition: WorkflowTask.cpp:133
unsigned long getBytesWritten() const
Get the number of bytes written by the task.
Definition: WorkflowTask.cpp:436
double task_failed
Task's failed time.
Definition: WorkflowTask.h:222
unsigned long getNumCoresAllocated()
Returns the number of cores allocated for this task's most recent execution or 0 if an execution atte...
Definition: WorkflowTask.cpp:764
Workflow * getWorkflow() const
Get the workflow that contains the task.
Definition: WorkflowTask.cpp:277
void setState(WorkflowTask::State)
Set the visible state of the task.
Definition: WorkflowTask.cpp:296
double computation_end
Task's computation end time.
Definition: WorkflowTask.h:214
double getWriteOutputEndDate()
Get the task's most recent write output end date.
Definition: WorkflowTask.cpp:700
A computational task in a Workflow.
Definition: WorkflowTask.h:27
double getEndDate()
Get the task's most recent end date.
Definition: WorkflowTask.cpp:652
double computation_start
Task's computation start time.
Definition: WorkflowTask.h:212
unsigned int getFailureCount()
Get the number of times a task has failed.
Definition: WorkflowTask.cpp:603
double getMemoryRequirement() const
Get the memory requirement of the task.
Definition: WorkflowTask.cpp:160
A data structure that keeps track of a task's execution event times.
Definition: WorkflowTask.h:204
WorkflowTask::State getUpcomingState() const
Get the state of the task.
Definition: WorkflowTask.cpp:214
Abstraction of a job used for executing tasks in a Workflow.
Definition: WorkflowJob.h:34
void setBytesRead(unsigned long)
Set the number of bytes read by the task.
Definition: WorkflowTask.cpp:428
double getFailureDate()
Get the task's most recent failure date.
Definition: WorkflowTask.cpp:708
std::string getExecutionHost()
Returns the name of the host on which the task has most recently been executed, or "" if the task has...
Definition: WorkflowTask.cpp:756
void setStartDate(double date)
Set the task's start date.
Definition: WorkflowTask.cpp:453
WorkflowTaskExecution(double task_start)
Constructor.
Definition: WorkflowTask.h:236
void setJob(WorkflowJob *job)
Set the task's containing job.
Definition: WorkflowTask.cpp:355
unsigned long getPriority() const
Get the task priority. By default, priority is 0.
Definition: WorkflowTask.cpp:388
unsigned long getTopLevel()
Returns the task's top level (max number of hops on a reverse path up to an entry task....
Definition: WorkflowTask.cpp:747
std::string getClusterID() const
Get the cluster Id for the task.
Definition: WorkflowTask.cpp:371
unsigned long getBytesRead() const
Get the number of bytes read by the task.
Definition: WorkflowTask.cpp:420
InternalState
Task state enum.
Definition: WorkflowTask.h:151
@ NOT_READY
Not ready (parents have not completed)
Definition: WorkflowTask.h:64
double getComputationStartDate()
Get the tasks's most recent computation start date.
Definition: WorkflowTask.cpp:660
void setEndDate(double date)
Set the task's end date.
Definition: WorkflowTask.cpp:463
void setWriteOutputEndDate(double date)
Set the date when the write output portion of a WorkflowTask has completed.
Definition: WorkflowTask.cpp:552
std::stack< WorkflowTaskExecution > getExecutionHistory()
Get the execution history of this task.
Definition: WorkflowTask.cpp:594
void setInternalState(WorkflowTask::InternalState)
Set the internal state of the task.
Definition: WorkflowTask.cpp:286
double getReadInputStartDate()
Get the task's most recent read input start date.
Definition: WorkflowTask.cpp:676
std::vector< WorkflowTask * > getParents() const
Get the parents of a task.
Definition: WorkflowTask.cpp:196
void setReadInputStartDate(double date)
Set the date when the read input portion of a WorkflowTask has begun.
Definition: WorkflowTask.cpp:507
void setAverageCPU(double)
Set the task average CPU usage.
Definition: WorkflowTask.cpp:412
void addOutputFile(WorkflowFile *file)
Add an output file to the task.
Definition: WorkflowTask.cpp:85
WorkflowJob * getJob() const
Get the task's containing job.
Definition: WorkflowTask.cpp:363
double task_start
Task's start time.
Definition: WorkflowTask.h:206
unsigned long getNumberOfParents() const
Get the number of parents of a task.
Definition: WorkflowTask.cpp:187
void setWriteOutputStartDate(double date)
Set the date when the write output portion of a WorkflowTask has begun.
Definition: WorkflowTask.cpp:537
A data file used/produced by a WorkflowTask in a Workflow.
Definition: WorkflowFile.h:26
double getFlops() const
Get the number of flops of the task.
Definition: WorkflowTask.cpp:124
void setColor(std::string)
Set the task's color.
Definition: WorkflowTask.cpp:806
std::vector< WorkflowFile * > getInputFiles()
Get the list of input WorkflowFile objects for the task.
Definition: WorkflowTask.cpp:618
double getReadInputEndDate()
Get the task's most recent read input end date.
Definition: WorkflowTask.cpp:684
double task_terminated
Task's terminated time.
Definition: WorkflowTask.h:224