• Type: Bug
• Resolution: Fixed
• Priority: Major
• Component/s: durable-task-plugin
• Labels:

  None
• Environment:

  z/OS

[JENKINS-72436] sh step on z/OS with encoding parameter set to UTF-8 hangs when trying to read jenkins-result.txt

Lutz Neugebauer created issue - 2023-12-19 09:25

Lutz Neugebauer made changes - 2023-12-19 09:28

Description

Original: When running sh step on z/OS with encoding set to UTF-8 the step execution hangs when trying to read the jenkins-result.txt file. This is because the file is in native EBCDIC encoding, but it is read with UTF-8 encoding specified in sh step. Following pipeline is used for reproduction:   {code:java} pipeline {     agent{node('zOS-node')}    stages {         stage('example for bugreport')   {             steps {                 sh(script:'''#!/bin/sh -e                     echo "UTF-8 output" | iconv -f 1047 -t utf-8''',                      encoding:'UTF-8')             }         }     } } {code} Execution results in: {code:java} [Pipeline] // stage [Pipeline] withEnv [Pipeline] { [Pipeline] stage [Pipeline] { (example for bugreport) [Pipeline] sh UTF-8 output Cannot contact zOS-node: java.io.IOException: corrupted content in /[path removed]/durable-040db3a4/jenkins-result.txt: java.lang.NumberFormatException: For input string: "�" {code} In my understanding on z/OS the jenkins-result.txt is always in native encoding as it is generated via shell scripts generated at [https://github.com/jenkinsci/durable-task-plugin/blob/aa22cf15d5e07bd99247e71ed467ae5107b74d73/src/main/java/org/jenkinsci/plugins/durabletask/BourneShellScript.java#L231] and following lines. Hence, it should be read with native encoding instead of encoding specified in sh step (which is intended for output/payload). I’ll come up with a PR.

New: When running sh step on z/OS with encoding set to UTF-8 the step execution hangs when trying to read the jenkins-result.txt file. This is because the file is in native EBCDIC encoding, but it is read with UTF-8 encoding specified in sh step. Following pipeline is used for reproduction:   {code:java} pipeline {     agent{node('zOS-node')}        stages {         stage('example for bugreport')   {             steps {                 sh(script:'''#!/bin/sh -e                     echo "UTF-8 output" | iconv -f 1047 -t utf-8''',                      encoding:'UTF-8')             }         }     } } {code} Execution results in: {code:java} [Pipeline] // stage [Pipeline] withEnv [Pipeline] { [Pipeline] stage [Pipeline] { (example for bugreport) [Pipeline] sh UTF-8 output Cannot contact zOS-node: java.io.IOException: corrupted content in /[path removed]/durable-040db3a4/jenkins-result.txt: java.lang.NumberFormatException: For input string: "�" {code} In my understanding on z/OS the jenkins-result.txt is always in native encoding as it is generated via shell scripts generated at [https://github.com/jenkinsci/durable-task-plugin/blob/aa22cf15d5e07bd99247e71ed467ae5107b74d73/src/main/java/org/jenkinsci/plugins/durabletask/BourneShellScript.java#L231] and following lines. Hence, it should be read with native encoding instead of encoding specified in sh step (which is intended for output/payload). I’ll come up with a PR.

Collapse comment: Lutz Neugebauer added a comment - 2023-12-20 15:37

Lutz Neugebauer added a comment - 2023-12-20 15:37

I raised PR#199.

Expand comment: Lutz Neugebauer added a comment - 2023-12-20 15:37

Lutz Neugebauer added a comment - 2023-12-20 15:37 I raised PR#199 .

Lutz Neugebauer made changes - 2024-01-05 07:57

Link

New: This issue is related to JENKINS-70660 [ JENKINS-70660 ]

Lutz Neugebauer made changes - 2024-01-14 20:07

Released As		New: https://github.com/jenkinsci/durable-task-plugin/releases/tag/543.v262f6a_803410
Resolution		New: Fixed [ 1 ]
Status	Original: Open [ 1 ]	New: Resolved [ 5 ]