有的情况下，用户可能不能访问Dr.Elephant的UI界面，但是需要和Dr.Elephant进行交互来获得Dr.Elephant的一些分析结果。Dr.Elephant提供了Rest API来满足用户的这个需求。
根据ID获得应用信息
根据任务的ID，可以获得任务的详细信息以及启发式算法的分析结果。
URI
http://<dr-elephant-host:port>/rest/job
请求参数
|
parameter |
description |
|
id |
The app or job id to search |
返回体
|
parameter |
description |
|
id |
The application id |
|
name |
The name of the application |
|
username |
User who submitted the application |
|
queueName |
The queue the application was submitted to |
|
startTime |
The time in which application started |
|
finishTime |
The time in which application finished |
|
trackingUrl |
The Job history page of the app |
|
jobType |
Type of the job e.g. pig |
|
severity |
Aggregate severity of all the heuristics. Ranges from 0(LOW) to 4(CRITICAL) |
|
score |
The application score which is the sum of heuristic scores |
|
workFlowDepth |
The application depth in the scheduled flow. Depth starts from 0 |
|
scheduler |
The scheduler which triggered the application |
|
jobName |
The name of the job in the flow to which this app belongs |
|
jobExecId |
A unique reference to a specific execution of the job/action(job in the workflow). |
|
flowExecId |
A unique reference to a specific flow execution. |
|
jobDefId |
A unique reference to the job in the entire flow independent of the execution. |
|
flowDefId |
A unique reference to the entire flow independent of any execution. |
|
jobExecUrl |
A url to the job execution on the schedule |
|
flowExecUrl |
A url to the flow execution on the scheduler |
|
jobDefUrl |
A url to the job definition on the scheduler |
|
flowDefUrl |
A url to the flow definition on the scheduler |
|
yarnAppHeuristicResults |
Detailed results on individual heuristics |
根据执行ID获得应用信息
根据任务的执行ID或者Azkaban的执行URL，可以获得所有的任务信息。
URI
http://<dr-elephant-host:port>/rest/jobexec
请求参数
|
parameter |
description |
|
id |
The job execution id or job execution url(azkaban) to search |
返回结果
返回结果包含了这个任务的所有子任务和作业。这个返回结果序列表格可以参考上面的“根据ID获得应用信息”部分
根据任务流ID来获取应用
根据任务流的ID或者Azkaban的任务执行url可以获得任务流的详细信息。
URI
http://<dr-elephant-host:port>/rest/flowexec
请求参数
|
parameter |
description |
|
id |
The flow execution id or flow execution url(azkaban) to search |
返回结果
返回结果包含了这个任务流的所有子任务和作业。这个返回结果序列表格可以参考上面的“根据ID获得应用信息”部分
搜索任务
根据一定的参数可以搜索任务和任务流。Rest API支持通过所有的参数来进行搜索。
URI
http://<dr-elephant-host:port>/rest/search
请求参数
|
parameter |
description |
|
id |
The job/app execution id to search. If this is provided, no other parameter is valid |
|
flow-exec-id |
The flow execution id or url(azkaban) to search, if this is provided, no other parameter is valid |
|
username |
To filter jobs by the user |
|
severity |
To filter all jobs based on severity |
|
job-type |
The type of jobs to search |
|
analysis |
The heuristic name. This goes together with severity. |
|
finished-time-begin |
The start range of finish time. (Milliseconds since epoch) |
|
finished-time-end |
The end range of finish time. (Milliseconds since epoch) |
|
started-time-begin |
The start range of start time. (Milliseconds since epoch) |
|
started-time-end |
The end range of start time. (Milliseconds since epoch) |
返回结果
返回结果包含了这个任务流的所有任务。这个返回结果序列表格可以参考上面的“根据ID获得应用信息”部分
比较任务的执行
给定两个任务的ID，会返回两个任务比较的结果
URI
http://<dr-elephant-host:port>/rest/compare
请求参数
|
parameter |
description |
|
flow-exec-id1 |
Flow execution id or url(azkaban) of first flow |
|
flow-exec-id2 |
Flow execution id or url(azkaban) of second flow |
返回结果
结果的格式如下：
{
job_def_id1: {
flow_exec_id1 : {
app1_details,
app2_details
},
flow_exec_id2 : {
app1_details,
app2_details
}
},
job_def_id2: {
flow_exec_id1 : {
app1_details,
app2_details,
app3_details
},
flow_exec_id2 : {
app1_details,
app2_details
app3_details
}
}
}
app1_details的结果是应用的详细返回结果，这个返回结果序列表格可以参考上面的“根据ID获得应用信息”部分
作者简介：屈世超，对高并发系统设计开发感兴趣，现专注于大数据开发工作。曾任职小米科技公司服务端后台开发工程师，现担任EverString数据平台组高级开发工程师。
