You are viewing a plain text version of this content. The canonical link for it is here.
Posted to hdfs-dev@hadoop.apache.org by "Siddharth Wagle (JIRA)" <ji...@apache.org> on 2019/05/21 22:17:00 UTC
[jira] [Created] (HDDS-1575) Add support for storage policies to
Pipelines
Siddharth Wagle created HDDS-1575:
-------------------------------------
Summary: Add support for storage policies to Pipelines
Key: HDDS-1575
URL: https://issues.apache.org/jira/browse/HDDS-1575
Project: Hadoop Distributed Data Store
Issue Type: Sub-task
Components: SCM
Reporter: Siddharth Wagle
The Pipeline level storage policy can be thought of segregating write bandwidth for high throughput writes or random reads use cases when SSDs, NVMe or RAM disks are involved by not having to figure out write pipelines on the fly. The Datanode would need to read the pipeline policy and route write to appropriate storage volumes. An Ozone mover can be provided for the archival data use case.
Additionally, it makes sense to permeate the policy information to Containers, this will help when allocateBlock call for a key with storage policy set to ALL_SSD, to be routed to an open container on a Pipeline with the same policy setting. The caveat is that at least one open pipeline needs to be maintained at all times in order to support storage the policy setting.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-dev-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-dev-help@hadoop.apache.org