[ https://issues.apache.org/jira/browse/HDFS-8476?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14560140#comment-14560140 ]
Xiaoyu Yao commented on HDFS-8476: ---------------------------------- [~tongshiquan], thanks for reporting the issue. I tried the scenario with a single node cluster with a 150MB file and slightly smaller quota size 400MB but cannot repro it. {code} $ hdfs dfs -mkdir /HOT $ hdfs dfs -put hadoop-3.0.0-SNAPSHOT.tar.gz /HOT/FILE1 $ hdfs dfsadmin -setSpaceQuota 400M -storageType DISK /HOT $ hdfs storagepolicies -setStoragePolicy -path /HOT -policy HOT Set storage policy HOT on /HOT $ hdfs dfs -count -q -h -v -t DISK /HOT DISK_QUOTA REM_DISK_QUOTA PATHNAME 400 M 256.6 M /HOT $ hdfs dfs -put hadoop-3.0.0-SNAPSHOT.tar.gz /HOT/FILE2 $ hdfs dfs -count -q -h -v -t DISK /HOT DISK_QUOTA REM_DISK_QUOTA PATHNAME 400 M 113.3 M /HOT $ hdfs dfs -put hadoop-3.0.0-SNAPSHOT.tar.gz /HOT/FILE3 put: Quota by storage type : DISK on path : /HOT is exceeded. quota = 400 MB but space consumed = 414.71 MB {code} > quota can't limit the file which put before setting the storage policy > ---------------------------------------------------------------------- > > Key: HDFS-8476 > URL: https://issues.apache.org/jira/browse/HDFS-8476 > Project: Hadoop HDFS > Issue Type: Bug > Components: HDFS > Affects Versions: 2.7.0 > Reporter: tongshiquan > Assignee: kanaka kumar avvaru > Priority: Minor > Labels: QBST > > test steps: > 1. hdfs dfs -mkdir /HOT > 2. hdfs dfs -put 1G.txt /HOT/file1 > 3. hdfs dfsadmin -setSpaceQuota 6442450944 -storageType DISK /HOT > 4. hdfs storagepolicies -setStoragePolicy -path /HOT -policy HOT > 5. hdfs dfs -put 1G.txt /HOT/file2 > 6. hdfs dfs -put 1G.txt /HOT/file3 > 7. hdfs dfs -count -q -h -v -t DISK /HOT > In step6 file put should fail, because /HOT/file1 and /HOT/file2 have reach > the directory /HOT space quota 6GB (1G*3 replicas + 1G*3 replicas), but here > it success, and in step7 count shows remaining quota is -3GB > FYI, if change the turn of step3 and step4, then it turns out normal -- This message was sent by Atlassian JIRA (v6.3.4#6332)