[
https://issues.apache.org/jira/browse/HBASE-7365?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13534469#comment-13534469
]
stack commented on HBASE-7365:
------------------------------
Makes sense doing stuff in .tmp (which mirrors table .tmp and region .tmp).
In below... the archive tool will be able to make sense of the deleted table?
+ if (fs.exists(tmpdir)) {
+ // Archive table in temp, maybe are failed deletion left over,
+ // if not the cleaner will take care of them.
+ for (Path tabledir: FSUtils.getTableDirs(fs, tmpdir)) {
+ for (Path regiondir: FSUtils.getRegionDirs(fs, tabledir)) {
+ HFileArchiver.archiveRegion(fs, this.rootdir, tabledir, regiondir);
+ }
+ }
+ fs.delete(tmpdir, true);
Do you want to check the returned value on the above delete?
Why not call checkTempDir rather than do this?
+ // Ensure temp exists
+ if (!fs.exists(tempdir) && !fs.mkdirs(tempdir)) {
+ throw new IOException("HBase temp directory '" + tempdir + "' creation
failure.");
I love it when TODOs get cleaned up...
- // TODO: Currently we make the table descriptor and as side-effect the
- // tableDir is created. Should we change below method to be createTable
- // where we create table in tmp dir with its table descriptor file and then
- // do rename to move it into place?
Should we remove the old handleCreateTable now we have your fancy new one? Or
is it still used?
Patch looks good otherwise Matteo.
> Safer table creation and deletion using .tmp dir
> ------------------------------------------------
>
> Key: HBASE-7365
> URL: https://issues.apache.org/jira/browse/HBASE-7365
> Project: HBase
> Issue Type: Improvement
> Components: master
> Reporter: Matteo Bertozzi
> Assignee: Matteo Bertozzi
> Fix For: 0.96.0
>
> Attachments: HBASE-7365-v0.patch
>
>
> Currently tables are created in the root directory, and the removal works on
> the root directory.
> Change the code to use a /hbase/.tmp directory to make the creation and
> removal a bit safer
> Table Creation steps
> * Create the table descriptor (table folder, in /hbase/.tmp/)
> * Create the table regions (always in temp)
> * Move the table from temp to the root folder
> * Add the regions to meta
> * Trigger assignment
> * Set enable flag in ZooKeeper
> Table Deletion steps
> * Wait for regions in transition
> * Remove regions from meta (use bulk delete)
> * Move the table in /hbase/.tmp
> * Remove the table from the descriptor cache
> * Remove table from zookeeper
> * Archive the table
> The main changes in the current code are:
> * Writing to /hbase/.tmp and then rename
> * using bulk delete in DeletionTableHandler
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
