cube building with huge memory, for 5000 records 2.3GB

[suresh m <su...@gmail.com>]

Hi,

My kylin cube taking huge memory just for few records. Even i changed
compression property as Snappy but still occupying huge memory.

Attaching screenshot for your reference.

Suggestions are welcome to reduce cube size.

Regards,
Suresh

Re: cube building with huge memory, for 5000 records 2.3GB

[ShaoFeng Shi <sh...@apache.org>]

You need check your Hadoop configuration; Kylin uses your default
configuration to request resource from Yarn; If you observed it uses huge
mount of mem, it may indicate the configuration is incorrect or the cube
design got problem.

Without detail information we couldn't judge what's wrong in your side.
Suggest you do more investigation. Usually the dimension in a cube should
be controlled < 15; if beyond that you need divide them into multiple
aggregation groups.

2017-04-28 18:13 GMT+08:00 suresh m <su...@gmail.com>:

> Hi,
>
> My kylin cube taking huge memory just for few records. Even i changed
> compression property as Snappy but still occupying huge memory.
>
> Attaching screenshot for your reference.
>
> Suggestions are welcome to reduce cube size.
>
> Regards,
> Suresh
>
>
>



-- 
Best regards,

Shaofeng Shi 史少锋