[
https://issues.apache.org/jira/browse/HDFS-2832?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13813539#comment-13813539
]
Arpit Agarwal commented on HDFS-2832:
-------------------------------------
Konstantin,
{quote}
I thought that the storageID approach that each DN is able to generate a unique
id independently of the others is a good feature to retain.
{quote}
Storage (UU)IDs are independently generated on the Datanode in
{{DataStorage#format}}.
{quote}
UUID as you noted is not unique and needs to be coordinated through NameNode.
{quote}
Not true. {{UUID#randomUUID}} generates RFC-4122 compliant UUIDs which are
unique for all practical purposes without NameNode coordination.
{quote}
You can also add to storageID an attribute that characterizes the disk volume
or the directory as a new component. Examples of the new attribute could be
disk serial number, or the storage directory inode number. It seems that
introduction of UUIDs was unnecessary, unless of course I missed some context.
{quote}
Part of the rationale is in HDFS-5115. Making them UUIDs simplifies the
generation logic. Decoupling them from volume/directory characteristics allows
future storage media that do not have a disk serial number or inode number.
> Enable support for heterogeneous storages in HDFS
> -------------------------------------------------
>
> Key: HDFS-2832
> URL: https://issues.apache.org/jira/browse/HDFS-2832
> Project: Hadoop HDFS
> Issue Type: New Feature
> Affects Versions: 0.24.0
> Reporter: Suresh Srinivas
> Assignee: Suresh Srinivas
> Attachments: 20130813-HeterogeneousStorage.pdf, h2832_20131023.patch,
> h2832_20131023b.patch, h2832_20131025.patch, h2832_20131028.patch,
> h2832_20131028b.patch, h2832_20131029.patch, h2832_20131103.patch,
> h2832_20131104.patch
>
>
> HDFS currently supports configuration where storages are a list of
> directories. Typically each of these directories correspond to a volume with
> its own file system. All these directories are homogeneous and therefore
> identified as a single storage at the namenode. I propose, change to the
> current model where Datanode * is a * storage, to Datanode * is a collection
> * of strorages.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
