[ https://issues.apache.org/jira/browse/HDFS-728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12769414#action_12769414 ]
Konstantin Boudnik commented on HDFS-728: ----------------------------------------- Right, I've understood this. My question is: "How do we suppose to run it in Hudson then?" It has to be executed by Hudson on a regular basis if it suppose to be a part of standard development cycle. Shell it be moved to a special load (or stressful) test suite? It might be an open discussion for this. However, there's no way to run a command line application as a part of CI build/test cycle. > Creat a comprehensive functional test for append > ------------------------------------------------ > > Key: HDFS-728 > URL: https://issues.apache.org/jira/browse/HDFS-728 > Project: Hadoop HDFS > Issue Type: New Feature > Affects Versions: 0.21.0 > Reporter: Hairong Kuang > Assignee: Hairong Kuang > Fix For: 0.21.0 > > Attachments: appendTest.patch > > > This test aims to do > 1. create a file of len1; > 2. reopen the file for append; > 3. write len2 bytes to the file and hflush; > 4. write len3 bytes to the file and close the file; > 5. validate the content of the file. > Len1 ranges from [0, 2*BLOCK_SIZE+1], len2 ranges from [0, BLOCK_SIZE+1], and > len3 ranges from [0, BLOCK_SIZE+1]. The test tries all combination of len1, > len2, and len3. To minimize the running time, bytes per checksum is set to be > 4 bytes, each packet size is set to be bytes per checksum, and each block > contains 2 packets. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.