hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "stack (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-14123) HBase Backup/Restore Phase 2
Date Wed, 16 Nov 2016 06:56:58 GMT

    [ https://issues.apache.org/jira/browse/HBASE-14123?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15669647#comment-15669647
] 

stack commented on HBASE-14123:
-------------------------------

I tried it. Here are some notes.

I applied patch and restarted. Nothing in logs about a backup table. Good.

I tried this: $ ./hbase/bin/hbase --config ~/conf_hbase/ backup

... and got this message:

Backup is not enabled. To enable backup, set 'hbase.backup.enabled'=true and restart the cluster

... which is reasonable (might want to fix the single quotes above to include the =true.

I ran ./hbase/bin/hbase --config ~/conf_hbase/ backup and got nice listing of commands.

I notice that if I run backup command without arg., it does right thing printing out the help.

  backup_root     Full path to store the backup image needs an example... is the spec for
path a full hdfs url or a local reference?

I tried local reference and got:

ERROR: invalid backup destination: /User/stack/Downloads/bkup

Providing full on hdfs URL seems to do the job.

stack@ve0524:~$ ./hbase/bin/hbase --config ~/conf_hbase/ backup create full hdfs://ve0524.halxg.cloudera.com:8020/user/stack/backup
ycsb

Can we file an issue to do a better output when doing history?

{code}
ID             : backup_1479277134105
Type           : FULL
Tables         : x_1
State          : COMPLETE
Start time     : Tue Nov 15 22:18:54 PST 2016
End time       : Tue Nov 15 22:19:11 PST 2016
Progress       : 100

ID             : backup_1479274617204
Type           : FULL
Tables         : ycsb
State          : COMPLETE
Start time     : Tue Nov 15 21:36:57 PST 2016
End time       : Tue Nov 15 21:52:48 PST 2016
Progress       : 100

ID             : backup_1479273680731
Type           : FULL
Tables         : ycsb
State          : RUNNING
Start time     : Tue Nov 15 21:21:21 PST 2016
Phase          : null
Progress       : 0
{code}

A single line would be good when lots of backups. The line as JSON would be parseable/actionable
by other tools. Can fix later.

Should 'history' be 'list'? History is not 'complete' as I'd expect after doing a few deletes.
Can fix later.

History shows all about a backup. 'describe' shows all about a single backup.  Maybe 'list'
shows all and 'list' with a backup id shows the ids specified only? Can be done later. 'history'
seems odd here.

'progress' help doesn't say it requires a backup id.

So, given the above note on how you format the backup emissions, it is interesting that the
set output has this form, a new format completely:

{code}
x={ycsb,x_1}
{code}

Would be good to unify the output formats here or at least have them mildly related?

Formatting is off here:

{code}
stack@ve0524:~$ ./hbase/bin/hbase --config ~/conf_hbase/ restore
SLF4J: Class path contains multiple SLF4J bindings.
SLF4J: Found binding in [jar:file:/home/stack/hbase-2.0.0-SNAPSHOT/lib/slf4j-log4j12-1.6.1.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: Found binding in [jar:file:/home/stack/hadoop-2.7.3-SNAPSHOT/share/hadoop/common/lib/slf4j-log4j12-1.7.10.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.
SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory]
2016-11-15 22:40:09,081 DEBUG [main] backup.RestoreDriver: Will automatically restore all
the dependencies
Usage: bin/hbase restore <backup_path> <backup_id> <table(s)> [options]
  backup_path     Path to a backup destination root
  backup_id       Backup image ID to restore  table(s)        Comma-separated list of tables
to restore
{code}

Restore like backup needs defaults and example args.

Please file subissues to fix the above but so far no blocker for merge. Let me do a bit more
in the morning. Running a restore now...  If things basically work and review checks out,
we can merge and pick up the above small stuff subsequently.... Will be back.













> HBase Backup/Restore Phase 2
> ----------------------------
>
>                 Key: HBASE-14123
>                 URL: https://issues.apache.org/jira/browse/HBASE-14123
>             Project: HBase
>          Issue Type: Umbrella
>            Reporter: Vladimir Rodionov
>            Assignee: Vladimir Rodionov
>            Priority: Blocker
>             Fix For: 2.0.0
>
>         Attachments: 14123-master.v14.txt, 14123-master.v15.txt, 14123-master.v16.txt,
14123-master.v17.txt, 14123-master.v18.txt, 14123-master.v19.txt, 14123-master.v2.txt, 14123-master.v20.txt,
14123-master.v21.txt, 14123-master.v24.txt, 14123-master.v25.txt, 14123-master.v27.txt, 14123-master.v28.txt,
14123-master.v29.full.txt, 14123-master.v3.txt, 14123-master.v30.txt, 14123-master.v31.txt,
14123-master.v32.txt, 14123-master.v33.txt, 14123-master.v34.txt, 14123-master.v35.txt, 14123-master.v36.txt,
14123-master.v37.txt, 14123-master.v5.txt, 14123-master.v6.txt, 14123-master.v7.txt, 14123-master.v8.txt,
14123-master.v9.txt, 14123-v14.txt, HBASE-14123-for-7912-v1.patch, HBASE-14123-for-7912-v6.patch,
HBASE-14123-v1.patch, HBASE-14123-v10.patch, HBASE-14123-v11.patch, HBASE-14123-v12.patch,
HBASE-14123-v13.patch, HBASE-14123-v15.patch, HBASE-14123-v16.patch, HBASE-14123-v2.patch,
HBASE-14123-v3.patch, HBASE-14123-v4.patch, HBASE-14123-v5.patch, HBASE-14123-v6.patch, HBASE-14123-v7.patch,
HBASE-14123-v9.patch
>
>
> Phase 2 umbrella JIRA. See HBASE-7912 for design document and description. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message