Reply
Highlighted
Contributor
Posts: 51
Registered: ‎09-14-2017

CDH Filesystem layout recommendations

Hi, I am looking for some recommendations for a typical CDH cluster filesystem layout. Can someone provide info on how the filesystems should be setup in the Util, Edge, Master and Data nodes for the various components like OS, logs, data etc.  Thanks.

Contributor
Posts: 25
Registered: ‎02-23-2018

Re: CDH Filesystem layout recommendations

Hi @ebeb,

 

Some questions you should answer:

  • Do you know the data volumen?
  • Do you need for batch processing or real time?
  • Do you know the users volumen?
  • Do you know if you need HA?
  • Do yo know .....

 

Could you give more details of the scenario?

 

Regards,

Manu.

 

 

Contributor
Posts: 51
Registered: ‎09-14-2017

Re: CDH Filesystem layout recommendations

Assuming a very vanilla installation with average workload say 2 util/edge nodes, 3 master nodes, 10-20 data nodes. Data volume say 100-200TB, batch/realtime:50:50, 200-500users, yes HA needed. I am looking if someone can send me a df -h on their filesystem which was based on Cloudera recomendation. How the util,edge,master,datanode filesystems look like for a well-designed CDH cluster. 


@manuroman wrote:

Hi @ebeb,

 

Some questions you should answer:

  • Do you know the data volumen?
  • Do you need for batch processing or real time?
  • Do you know the users volumen?
  • Do you know if you need HA?
  • Do yo know .....

 

Could you give more details of the scenario?

 

Regards,

Manu.

 

 


 

Announcements