[ 
https://issues.apache.org/jira/browse/HDFS-7584?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14288118#comment-14288118
 ] 

Xiaoyu Yao commented on HDFS-7584:
----------------------------------

bq. But then how should we coordinate type-based quotas and the "traditional 
space quota"? Is the latter the summation of quotas from all types? If I set my 
SSD quota as 1GB and DISK quota as 2GB, do I need to separately set my overall 
space quota to be 3GB?

We propose not to correlate storage type quota with legacy space quota. Legacy 
space quota lacks fine-grain control of storage usage and is left as-is for 
backward compatibility.  If you have 1GB SSD quota dn 2GB DISK quota, which 
will limit your maximum overall space usage at 3GB of space quota. You don't 
need to set legacy quota in this case. 

bq. This is where it would be helpful to exchange type quotas. Revising the 
example a little bit, let's assume I have 3 blocks of SSD quota and no DISK 
quota, and the cluster has no SSD storage and plenty of DISK storage. When 
trying to create a 1 block file with replication factor of 3, under ONE_SSD 
policy, the user would expect a success because he has 3 blocks of quota in the 
most expensive type. I understand setting "exchange rates" needs more 
discussion. Can we at least allow more expensive quotas to be used on less 
expensive storages?

The example above looks like a configuration that needs admin to either add 
more SSD storage or increase the DISK quota. By default, quota should be simple 
and strict. The exchange rate actually could be a guideline for admin setup 
quotas of different storage types. If automatic correlation for quota of 
different storage types is desired, it should be explicitly set though 
policies. 

> Enable Quota Support for Storage Types (SSD) 
> ---------------------------------------------
>
>                 Key: HDFS-7584
>                 URL: https://issues.apache.org/jira/browse/HDFS-7584
>             Project: Hadoop HDFS
>          Issue Type: New Feature
>          Components: datanode, namenode
>            Reporter: Xiaoyu Yao
>            Assignee: Xiaoyu Yao
>         Attachments: HDFS-7584 Quota by Storage Type - 01202015.pdf, 
> HDFS-7584.0.patch
>
>
> Phase II of the Heterogeneous storage features have completed by HDFS-6584. 
> This JIRA is opened to enable Quota support of different storage types in 
> terms of storage space usage. This is more important for certain storage 
> types such as SSD as it is precious and more performant. 
> As described in the design doc of HDFS-5682, we plan to add new 
> quotaByStorageType command and new name node RPC protocol for it. The quota 
> by storage type feature is applied to HDFS directory level similar to 
> traditional HDFS space quota. 



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

Reply via email to