[
https://issues.apache.org/jira/browse/ASTERIXDB-2841?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17296591#comment-17296591
]
ASF subversion and git services commented on ASTERIXDB-2841:
------------------------------------------------------------
Commit 4b3583211aa27922555a78993b79f04afa4d8bdb in asterixdb's branch
refs/heads/master from Michael Blow
[ https://gitbox.apache.org/repos/asf?p=asterixdb.git;h=4b35832 ]
[ASTERIXDB-2841][*DB][STO] Encode multiple-dataverse parts as subdirs on disk
- Multipart dataverse names are expressed on disk as a directory tree
- The first part is expressed normally, subsequent parts have a carat (^)
prepended
Change-Id: Idcfc45eb7f39153349a13d2baecb784244bdf177
Reviewed-on: https://asterix-gerrit.ics.uci.edu/c/asterixdb/+/10324
Reviewed-by: Michael Blow <[email protected]>
Tested-by: Michael Blow <[email protected]>
> Storage incompatibilities with created entities
> -----------------------------------------------
>
> Key: ASTERIXDB-2841
> URL: https://issues.apache.org/jira/browse/ASTERIXDB-2841
> Project: Apache AsterixDB
> Issue Type: Bug
> Components: *DB - AsterixDB, STO - Storage
> Affects Versions: 0.9.6
> Reporter: Michael Blow
> Assignee: Michael Blow
> Priority: Critical
> Fix For: 0.9.7
>
>
> Currently there are multiple ways in which created entities conflict with
> storage:
> * Most filesystems have 255 byte or character limit per path segment, for
> multipart dataverses we can greatly exceed this limit, and currently
> multipart dataverses are represented using canonical form into a single path
> element
> * Index names can be 255 bytes long, which result in corresponding files
> exceeding the path limit on most file systems
> * Dataverse / dataset / index names can contain characters which are illegal
> on the filesystem, e.g. \ and / on Windows, NUL (\u0000) on Linux
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
