[
https://issues.apache.org/jira/browse/IGNITE-15447?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Alexey Scherbakov updated IGNITE-15447:
---------------------------------------
Description:
We should measure the effectiveness of IO subsystem comparing to ignite 2.
The goal is to maximise network bandwidth and TPS for peer to peer invoke
operations, which are the basis for all ignite-3 protocols.
The basic test scenario is calling invoke from one node to another using some
configurable byte array payload and threads count.
This benchmark can be integrated in the product to test user's network
capabilities.
Ignite 2 already includes the similar tool [1]
[1] org.apache.ignite.mxbean.IgniteMXBean#runIoTest
was:
The following scenarios should be tested (at least):
# invoke performance on local node (send to self). Can be implemented as JMH
benchmark.
# invoke performance between 2 physical nodes (peer 2 peer). Measure TPS and
consumed bandwidth. Can be integrated in the product to test user's network
capabilities.
> Implement benchmarks for network layer
> --------------------------------------
>
> Key: IGNITE-15447
> URL: https://issues.apache.org/jira/browse/IGNITE-15447
> Project: Ignite
> Issue Type: Task
> Reporter: Alexey Scherbakov
> Priority: Major
> Labels: ignite-3
>
> We should measure the effectiveness of IO subsystem comparing to ignite 2.
> The goal is to maximise network bandwidth and TPS for peer to peer invoke
> operations, which are the basis for all ignite-3 protocols.
> The basic test scenario is calling invoke from one node to another using some
> configurable byte array payload and threads count.
> This benchmark can be integrated in the product to test user's network
> capabilities.
> Ignite 2 already includes the similar tool [1]
> [1] org.apache.ignite.mxbean.IgniteMXBean#runIoTest
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
