BookmarkSubscribeRSS Feed
naveenraj
Quartz | Level 8

I have a unix environment and have installed a non distribited LASR server. So i want to know what is the maximum size i can load to lasr server before reaching its threshold which is set at 80% 

6 REPLIES 6
Mark_sas
SAS Employee

There's a description of how to check out memory usage of the LASR server and loaded tables in the Monitoring Server Memory Use and Monitoring Table Memory Use sections of the SAS LASR Analytic Server Reference Guide at http://support.sas.com/documentation/cdl/en/inmsref/68736/HTML/default/viewer.htm#p050lknh5xepngn1s2...

 


Register today and join us virtually on June 16!
sasglobalforum.com | #SASGF

View now: on-demand content for SAS users

JuanS_OCS
Amethyst | Level 16

Hi,

 

besides the great tip from @Mark_sas, which provides you a great reporting on your LASR servers and tables, an additional remark:

 

when you requested SAS VA to SAS, mostly you received a questionaire/study with some questions, as:

- expected amount of heavy and light users on your system

- expected total amount of data to be loaded in memory

- expected biggest table to be loaded into memory

- and others

 

After this questionaire, SAS forwarded the information to the SAS Excelence Center, where they could come back to you with an initial sizing of your server, which includes amount of CPUs, RAM size, storage, and other architecture details.

 

If the information of this questionaire changes, dramatically, I would recommend to initiate the same procedure with your SAS account manager. Otherwise, you and your users might experience a performance decrease of the SAS VA server.

SimonWilliams
SAS Employee

Hi naveenraj ,

 

In addition to replies from @Mark_sas and @JuanS_OCS you may find watching this YouTube video by Mark Thomas (SAS) of help:

 

https://www.youtube.com/watch?v=z1GdCyAT7mI

 

Cheers, Simon

 


Register today and join us virtually on June 16!
sasglobalforum.com | #SASGF

View now: on-demand content for SAS users

naveenraj
Quartz | Level 8

no i have not found a proper answer yet. I checked with sas consultants and they are saying to calculate the free ram memory. But i am not sure that would provide answer to my question. 

SimonWilliams
SAS Employee

Hi Naveenraj,

 

Please see the attached picture which is a screen shot from the youtube clip i shared with you.

 

It breaks it down nicely between VA with non-distributed or distributed LASR.

 

 When you ask "So i want to know what is the maximum size i can load to lasr server before reaching its threshold which is set at 80% " has a couple of aspects to it:

 

* If you have no tables loaded into LASR then the largest table (by default) you can load in to VA is 80% of the total memory on the system e.g. if you have 100gb of RAM then you can load an 80gb table into RAM. Depending on what the type of analyses you are being executed and what other processes are running on the system this may be ok but if performance is poor you should consider increasing the amount of RAM ;

 

* If you have have 60gbs worth of tables loaded into a 100gb machine, then if you attempt to load in 30gb table LASR will report an error because you will have passed the threshold;

 

* If you want to limit the size of the data tables then you can do that by setting the tables limit option.

 

If you take a look at the visual analytics administration guide it talks about overcapacity of the system and the expected behaviour. I think that will help you.

 

--Simon

 


Capture4.JPG

Register today and join us virtually on June 16!
sasglobalforum.com | #SASGF

View now: on-demand content for SAS users

suga badge.PNGThe SAS Users Group for Administrators (SUGA) is open to all SAS administrators and architects who install, update, manage or maintain a SAS deployment. 

Join SUGA 

CLI in SAS Viya

Learn how to install the SAS Viya CLI and a few commands you may find useful in this video by SAS’ Darrell Barton.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 6 replies
  • 5006 views
  • 3 likes
  • 4 in conversation