[infinispan-issues] [JBoss JIRA] Commented: (ISPN-183) Test performance of DefaultConsistentHash
[
https://jira.jboss.org/jira/browse/ISPN-183?page=com.atlassian.jira.plugi...
]
Manik Surtani commented on ISPN-183:
------------------------------------
ConsistentHashPerfTest is a manually run test that can be used to test the scalability of
the consistent hash function, over varying cluster sizes.
One run I just did (not definitive, of course) generates these results:
With 1 cache(s), time to do 100000 lookups was 6 milliseconds
With 10 cache(s), time to do 100000 lookups was 6 milliseconds
With 100 cache(s), time to do 100000 lookups was 7 milliseconds
With 1000 cache(s), time to do 100000 lookups was 7 milliseconds
With 10000 cache(s), time to do 100000 lookups was 31 milliseconds
Test performance of DefaultConsistentHash
-----------------------------------------
Key: ISPN-183
URL: https://jira.jboss.org/jira/browse/ISPN-183
Project: Infinispan
Issue Type: Task
Components: Distributed Cache
Affects Versions: 4.0.0.BETA1
Reporter: Manik Surtani
Assignee: Manik Surtani
Fix For: 4.0.0.CR1, 4.0.0.GA
A simple test that creates a bunch of dummy addresses and attempts to locate a key. Need
to test how this degrades as more nodes are added to the CH.
The purpose of this is certain applications may need to generate keys such that the key
is *always* mapped to the local node. One approach is to loop this until the random key
is mapped by locating the key and checking if the local address is in the server list.
Need to see if this is (a) costly, and whether (b) a better mechanism can be provided
here.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira