将每个阶段的数据从 jenkins 管道推送到 influx db 时如何修复 java.io.NotSerializable 异常？

Question

我正在尝试将阶段数据从 Jenkins 管道推送到 influx 数据库，但出现如下问题

问题：在 jenkin 构建控制台输出中，在每个阶段之后，我看到：java.io.NotSerializableException：sun.net.www.protocol.http.HttpURLConnection

詹金斯管道 2.150.2 InfluxDB 1.6.2

任何建议都会有所帮助。我是这个话题的新手。

注意：@NonCPS注解我已经注释掉了。如果我取消注释，那么它只发送第一阶段数据并退出并且无法迭代具有 20 个阶段数据的循环。

//Maps for Field type columns
myDataField1 = [:]
myDataField2 = [:]
myDataField3 = [:]

//Maps for Custom Field measurements
myCustomDataFields1 = [:]
myCustomDataFields2 = [:]
myCustomDataFields3 = [:]

//Maps for Tag type columns
 myDataTag1 = [:]
 myDataTag2 = [:]
 myDataTag3 = [:]

//Maps for Custom Tag measurements
myCustomDataTags1 = [:]
myCustomDataTags2 = [:]
myCustomDataTags3 = [:]

//@NonCPS
def pushStageData() {

def url_string = "${JENKINS_URL}job/ENO_ENG_TP/job/R421/13/wfapi/describe"
def get = new URL(url_string).openConnection();

get.addRequestProperty ("User-Agent","Mozilla/4.0");
get.addRequestProperty("Authorization", "Basic dZXZvceDIwMTk=");

//fetching the contents of the endpoint URL
def jsonText = get.getInputStream().getText();
//converting the text into JSON object using 
JsonSlurperClassic 
def jsonObject = new JsonSlurperClassic().parseText(jsonText)

// Extracting the details of all the stages present in that particular build number
for (int i=0; i<jsonObject.stages.size()-1; i++){ //size-1 to ignore the post stage
//populating the field type columns of InfluxDB measurements and pushing them to the map called myDataField1
def size = jsonObject.stages.size()-1
myDataField1['result'] = jsonObject.stages[i].status
myDataField1['duration'] = 
jsonObject.stages[i].durationMillis
myDataField1['stage_name'] = jsonObject.stages[i].name

//populating the tag type columns of InfluxDB 
measurements and pushing them to the map called 
myDataTag1
myDataTag1['result_tag'] = jsonObject.stages[i].status
myDataTag1['stage_name_tag'] = jsonObject.stages[i].name

//assigning field type columns to the measurement called CustomData
myCustomDataFields1['CustomData'] = myDataField1
//assigning tag type columns to the measurement called CustomData
myCustomDataTags1['CustomData'] = myDataTag1

//Push the data into influx instance
try
{ step([$class: 'InfluxDbPublisher', target: 'jenkins_data', customPrefix: null, customDataMapTags: myCustomDataTags1]) }

catch (err)
{ println ("pushStagData exception: " + err) }
}
}

预期：我想将jenkins管道数据的各个阶段推送到influx db。

Answer 1

尝试使用 jenkins 的标准方法，而不是手动创建对象。您可以使用 httpRequest instead of URL. This should fix exception you are getting. You can also use readJson 而不是 JsonSlurperClassic（后者是可选的，因为经典的 slurper 实际上是可序列化的）。

Answer 2

您只需要将 get 显式设置为 null 以确保有 none 不可序列化的对象被实例化。

def jsonText = get.getInputStream().getText();
get = null;