[
https://issues.jboss.org/browse/ISPN-2862?page=com.atlassian.jira.plugin....
]
Pedro Ruivo updated ISPN-2862:
------------------------------
Description:
With the same goal as L1 Cache, it applies a different technique to reduce the
communication overhead. The self-tunning data placement detects which are the keys most
accessed by each node, modifies the ConsistentHash and triggers the State Transfer in
order to move the keys for that nodes.
The main difference for L1 Cache is the fact that L1 creates a new copy of the
<key,value> in the requestor node, that needs to be invalidated later. The Data
Placement, really moves the <key,value> to that requestor.
Both techniques can be enabled at the same time.
This module depends of ISPN-2861, namely the top-key module.
was:This is a smarter L1 implementation.
Integrate the Self-tuning Data Placement from CloudTM into
Infinispan
---------------------------------------------------------------------
Key: ISPN-2862
URL:
https://issues.jboss.org/browse/ISPN-2862
Project: Infinispan
Issue Type: Feature Request
Reporter: Mircea Markus
Assignee: Pedro Ruivo
Fix For: 5.3.0.Final
Attachments: autoplacer.pdf
With the same goal as L1 Cache, it applies a different technique to reduce the
communication overhead. The self-tunning data placement detects which are the keys most
accessed by each node, modifies the ConsistentHash and triggers the State Transfer in
order to move the keys for that nodes.
The main difference for L1 Cache is the fact that L1 creates a new copy of the
<key,value> in the requestor node, that needs to be invalidated later. The Data
Placement, really moves the <key,value> to that requestor.
Both techniques can be enabled at the same time.
This module depends of ISPN-2861, namely the top-key module.
--
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