You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@ignite.apache.org by 罗 辉 <lu...@zetyun.com> on 2017/07/27 08:12:55 UTC

Is it possible to make use of SSD in a Ignite cluster?

Hello guys:

       My cluster has some SSDs, and I want to make use of them in my Ignite applications.

1.Is there a way to use SSD when my cluster memory is running out to cache the data?

2.Is there a parameter that control intermediate result storage directory, like the one "spark.local.dir" in apache spark?

3.Is there a parameter or a policy that Ignite can evict some data from memory to SSD based on LRU when the memory is running out?


Any relative ideas is appreciated.

thanks
San.Luo

答复: Is it possible to make use of SSD in a Ignite cluster?

Posted by 罗 辉 <lu...@zetyun.com>.
Thank you ,I got it.

________________________________
发件人: Michael Cherkasov <mi...@gmail.com>
发送时间: 2017年7月27日 16:54:52
收件人: user@ignite.apache.org
主题: Re: Is it possible to make use of SSD in a Ignite cluster?

Hi San Luo.

>1.Is there a way to use SSD when my cluster memory is running out to cache the data?
Ignite 2.1 can do this for you:
https://apacheignite.readme.io/v2.1/docs/distributed-persistent-store
it will be release this week.
>2.Is there a parameter that control intermediate result storage directory, like the one "spark.local.dir" in apache spark?
Check PersistentStoreConfiguration there's a parameter for this where you can define dir to store data.
>3.Is there a parameter or a policy that Ignite can evict some data from memory to SSD based on LRU when the memory is running out?
https://apacheignite.readme.io/v2.1/docs/evictions

Thanks,
Mikhail.

2017-07-27 11:12 GMT+03:00 罗 辉 <lu...@zetyun.com>>:

Hello guys:

       My cluster has some SSDs, and I want to make use of them in my Ignite applications.

1.Is there a way to use SSD when my cluster memory is running out to cache the data?

2.Is there a parameter that control intermediate result storage directory, like the one "spark.local.dir" in apache spark?

3.Is there a parameter or a policy that Ignite can evict some data from memory to SSD based on LRU when the memory is running out?


Any relative ideas is appreciated.

thanks
San.Luo

Re: Is it possible to make use of SSD in a Ignite cluster?

Posted by Michael Cherkasov <mi...@gmail.com>.
Hi San Luo.

>1.Is there a way to use SSD when my cluster memory is running out to cache
the data?
Ignite 2.1 can do this for you:
https://apacheignite.readme.io/v2.1/docs/distributed-persistent-store
it will be release this week.
>2.Is there a parameter that control intermediate result storage directory,
like the one "spark.local.dir" in apache spark?
Check PersistentStoreConfiguration there's a parameter for this where you
can define dir to store data.
>3.Is there a parameter or a policy that Ignite can evict some data from
memory to SSD based on LRU when the memory is running out?
https://apacheignite.readme.io/v2.1/docs/evictions

Thanks,
Mikhail.

2017-07-27 11:12 GMT+03:00 罗 辉 <lu...@zetyun.com>:

> Hello guys:
>
>        My cluster has some SSDs, and I want to make use of them in my
> Ignite applications.
>
> 1.Is there a way to use SSD when my cluster memory is running out to cache
> the data?
>
> 2.Is there a parameter that control intermediate result storage directory,
> like the one "spark.local.dir" in apache spark?
>
> 3.Is there a parameter or a policy that Ignite can evict some data from
> memory to SSD based on LRU when the memory is running out?
>
>
> Any relative ideas is appreciated.
>
> thanks
> San.Luo
>
>
>