[ 
https://issues.apache.org/jira/browse/IGNITE-5930?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Ksenia Rybakova updated IGNITE-5930:
------------------------------------
    Summary: Long running transactions while load test with 1 or many memory 
policies, 14-30 caches and >=24 servers  (was: Long running transactions while 
load test with 1 memory policy, 14 caches and >=24 servers)

> Long running transactions while load test with 1 or many memory policies, 
> 14-30 caches and >=24 servers
> -------------------------------------------------------------------------------------------------------
>
>                 Key: IGNITE-5930
>                 URL: https://issues.apache.org/jira/browse/IGNITE-5930
>             Project: Ignite
>          Issue Type: Bug
>    Affects Versions: 2.1
>            Reporter: Ksenia Rybakova
>         Attachments: ignite-base-load-config.xml, run-load.properties, 
> run-load.xml
>
>
> 1) Load config:
> - CacheRandomOperation benchmark
> - Caches:
> atomic partitioned cache
> 2 tx partitioned caches
> 2 atomic partitioned onheap caches
> 2 atomic replicated onheap caches
> 2 tx partitioned onheap caches
> 2 tx replicated onheap caches
> 3 partitioned atomic indexed caches
> - 1 default memory policy is used
> - Preload amount: 500K-2.5M entries per cache
> - Key range: 1M
> - Persistent store is disabled
> - 24 server nodes or more (3 nodes per physical host), 16 client nodes
> Long running transactions occur after preloading and throughput dropps to 0. 
> Note, that with 16 server nodes transactions are ok.
> Complete configs are attached.
> 2) Another config that causes long running transactions:
> - CacheRandomOperation benchmark
> - 30 caches with different config
> - Separate memory policy is used for each cache
> - Preload amount: 500K-2.5M entries per cache
> - Key range: 1M
> - Persistent store is disabled
> - 24 server nodes or more (3 nodes per physical host), 16 client nodes
> Complete configs are attached.
> Note, that with 16 server nodes long running transactions are still there, 
> but throughput is much higher.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to