BookmarkSubscribeRSS Feed
hassan0707
Fluorite | Level 6

Hello,

 

There are times in our SAS VA enviroment that one of the users loads a very large table adn it slows down the whole system. I want to know if there is a way to find out the detail of that particular use in SAS VA like the session ID, username, process ID and then use that process ID to kill that SAS VA session. I am working on Linux system and have access to only the linux box.

 

Appreciate quick replies.

6 REPLIES 6
Anand_V
Ammonite | Level 13
Hi @hassan0707,

I am hoping you are talking about SAS VA on 9.4 here.
If you administer distributed VA setup, you can use the Grid Monitor Utility by launching it from server. Here is the link below with steps for starting up.
http://support.sas.com/documentation/cdl/en/inmsref/68736/HTML/default/viewer.htm#p18hwo14vdm4i6n13g...

Steps to Monitoring Jobs:
http://support.sas.com/documentation/cdl/en/inmsref/68736/HTML/default/viewer.htm#n0oyml6xxpmv0an161...

However this isn't available on the non-distributed mode. If this doesn't help then may be you can try to setup some limit on the size of the tables that users can load.

Link to the steps here, Pg#28 Section: How to Limit Import Size

http://support.sas.com/documentation/cdl/en/vaag/69958/PDF/default/vaag.pdf

Thanks!
hassan0707
Fluorite | Level 6

Hi @Anand_V ,

 

Yes, the system is on SAS 9.4 and distributed one. Setting limit on the table to be loaded is not desirable in my case as the users are required to use that table. The issue comes up when many users start using the system with the bigger table in the memory. I will check the solutions mentioned and let you know.

 

Thanks for quick response!!

hassan0707
Fluorite | Level 6

Hello @Anand_V ,

 

I tried the steps given in the links you have mentioned. I was able to run the command without any error. However I did not get the expected window with results mentioned in the links. For me the result window says 0 machines , but there are 4 machines in the distributed system. I have attached the screenshot. Could you please suggest what I might be missing?

Anand_V
Ammonite | Level 13
Hi @hassan0707,

Can you try it launching from the path - /opt/TkGrid/bin/gridmon.sh?

If your environment has custom path set for storing TKGrid Binaries, navigate to that path. <Tk-Grid-root-path>/bin/gridmon.sh

Thanks!
hassan0707
Fluorite | Level 6

Hello @Anand_V ,

 

I was able to run the tool after correcting some path related mistakes. However when I switched to see the jobs I could see the same two LASR server processes in the list which I can see in the SAS VA Administration web page too. What I am actually looking for is to find a way to see user session connected by SAS VA and stop/kill those sessions if their memory usage or other parameters breach a threshold. I have attached a screenshot.

 

Neverthless thanks a lot for sharing the information. Any solution/suggestion is most welcomed.

 

Regards

Hassan

Anand_V
Ammonite | Level 13
Hi @hassan0707,

For sure, I don't think there is anything as such pre-built in the system, that could fetch you these details. I did find this on SAS Communities a similar requirement.

https://communities.sas.com/t5/SAS-Visual-Analytics/How-do-I-determine-users-currently-connected-in-...

Other way I can think of is on server, you can grep the word "sasexe" to list all the SAS process that are currently running in order to get things like processID, Username etc. however it won't give you an idea of how much memory it has consumed over the distributed VA LASR Server as the data is split across worker nodes.

If you are able to find anything on this, would suggest to share on this post as it would helpful to others as well.

HTH, Thanks!

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

Tips for filtering data sources in SAS Visual Analytics

See how to use one filter for multiple data sources by mapping your data from SAS’ Alexandria McCall.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 6 replies
  • 1025 views
  • 1 like
  • 2 in conversation