[jira] [Updated] (HDDS-4094) Support byte-leve write in Freon HadoopFsGenerator
[ https://issues.apache.org/jira/browse/HDDS-4094?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ASF GitHub Bot updated HDDS-4094: - Labels: pull-request-available (was: ) > Support byte-leve write in Freon HadoopFsGenerator > -- > > Key: HDDS-4094 > URL: https://issues.apache.org/jira/browse/HDDS-4094 > Project: Hadoop Distributed Data Store > Issue Type: Improvement >Reporter: Marton Elek >Assignee: Marton Elek >Priority: Major > Labels: pull-request-available > > Teragen seems to use the byte level write method of FSDataOutputStream > (write(byte) is used instead of write(byte[], int, int)). > It seems to be a good idea to extend existing `ContentGenerator` of ozone to > support the write in smaller chunks to make it easier to reproduce > performance problems. > Note: statistics from FileSystem instance: > {code} > Closing file system instance: 1257412274 >write.call: 11066 >write.allTime: 215951 >hsync.call: 1 >hsync.allTime: 3 >hflush.call: 0 >hflush.allTime: 0 >close.call: 4 >close.allTime: 62 > {code} > This was a teragen test with 1GB data (and statistics from one container). > write method seems to be called multiple times which means smaller write > buffer. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: ozone-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: ozone-issues-h...@hadoop.apache.org
[jira] [Updated] (HDDS-4094) Support byte-leve write in Freon HadoopFsGenerator
[ https://issues.apache.org/jira/browse/HDDS-4094?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Attila Doroszlai updated HDDS-4094: --- Labels: (was: pull-request-available) > Support byte-leve write in Freon HadoopFsGenerator > -- > > Key: HDDS-4094 > URL: https://issues.apache.org/jira/browse/HDDS-4094 > Project: Hadoop Distributed Data Store > Issue Type: Improvement >Reporter: Marton Elek >Assignee: Marton Elek >Priority: Major > Fix For: 0.7.0 > > > Teragen seems to use the byte level write method of FSDataOutputStream > (write(byte) is used instead of write(byte[], int, int)). > It seems to be a good idea to extend existing `ContentGenerator` of ozone to > support the write in smaller chunks to make it easier to reproduce > performance problems. > Note: statistics from FileSystem instance: > {code} > Closing file system instance: 1257412274 >write.call: 11066 >write.allTime: 215951 >hsync.call: 1 >hsync.allTime: 3 >hflush.call: 0 >hflush.allTime: 0 >close.call: 4 >close.allTime: 62 > {code} > This was a teragen test with 1GB data (and statistics from one container). > write method seems to be called multiple times which means smaller write > buffer. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: ozone-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: ozone-issues-h...@hadoop.apache.org