BookmarkSubscribeRSS Feed
merrittr
Obsidian | Level 7

We have 11 user/devs that currently use Base SAS and curated SAS datasets (some) may use some ODBC connections for PROC SQL , what kind of timelines would one expect for this kind of migration?

To me it seem s trivial SAS consultants have the VA server up and we  just need to migrate devs and data sources? our internal IT resources say 1 year!!!

 

25 REPLIES 25
SASKiwi
PROC Star

What about their applications, data etc? It all depends on the number and size of the existing applications and data how long this will take. What SAS version and underlying OS are you migrating from and to? 

merrittr
Obsidian | Level 7
Sas 9.4 on window , no applications.

In sas va what role does the metadata server play if all your datasets are just sas data set
SASKiwi
PROC Star

Don't your SAS users develop SAS programs aka SAS applications and they need to be migrated? SAS VA on Viya or SAS VA on SAS 9.4? If the latter then ALL datasets have to be defined in SAS metadata. 

Quentin
Super User

When you say:


@merrittr wrote:
Sas 9.4 on window , no applications.



When you SAS 9.4 on Windows, do you mean you have 10-12 epidemiologists who are using Windows PC SAS?  And you're planning a migration to Viya? 

 

Or are they using  SAS 9.4 on a windows server now, and they're using EG or SAS Studio to submit code?

 

You might want to look into SAS analytics pro https://www.sas.com/en_us/software/analytics-pro.html if you've just got analysts who need to crunch data. 

 

Moving PC SAS users to become server users will not be trivial, both in terms of the resources needed to support the server, and training/learning curve for the SAS programmers.

 

BASUG is hosting free webinars Next up: Mike Raithel presenting on validating data files on Wednesday July 17. Register now at the Boston Area SAS Users Group event page: https://www.basug.org/events.
AhmedAl_Attar
Rhodochrosite | Level 12

Be aware SAS VA is a no/low code SAS platform.

All your data has to be loaded into memory before it can be used by VA (Visual Analytics)

So I would recommend you

  1. Analyze and categorize your dev's applications 
    • Data Engineering apps (ETL/ELT)
    • Reporting apps (Listings & Charts) 
  2. Check the features and capabilities of the SAS VA platform and try to see how they meet your apps needs
  3. Schedule required training for your users/devs & SAS Administrator(s), Because SAS VA is very different than Base/Foundation SAS   

Hope this helps

Ahmed

merrittr
Obsidian | Level 7
These are epidemiologist all there apps are statistical tables no etl.
merrittr
Obsidian | Level 7
What role does the va server meta data servers role in getting the data over and used
SASKiwi
PROC Star

I'm assuming you are using SAS VA on SAS 9.4 (not SAS Viya) - all SAS data loaded into VA LASR in-memory tables need to be defined in SAS metadata. To what extent do you use SAS metadata on your existing SAS installation? Do you have SAS libraries defined in metadata for example? Your current use of metadata will influence what you need to migrate and define on your new installation.

merrittr
Obsidian | Level 7

we currently use SAS Base 9.4 with flat sas datafiles produced for the Epidemiologist

we are moving to 

SAS VA on Viya (cloud) 

 

so to transition the data files that the epis have all have to be ingested into the metadata server?

Quentin
Super User

When you say:


@merrittr wrote:

we currently use SAS Base 9.4 


Where is SAS running currently?  Is it SAS on a remote server (which you connect to via Enterprise Guide or SAS Studio) or is it PC SAS running on your local PC.

 

If it's SAS on a server, you will want to get your IT group / server administrators involved in this assessment.  If your internal IT who said it will take a year are the folks who run your current environment, and will be responsible for leading the migration to Viya, then I would either trust them when they say it will take a year (at least), or question them about why it will take that long.  Perhaps your internal IT and/or the SAS service provider that you are planning to contract with can provide more explanation.

 

You might also ask your SAS service provider to give you access to a sandbox, so a couple of your epidemiologists can connect to it and play a bit, to see how workable it feels. If the epidemiologists are running PC SAS locally now, working on a remote server will be an adjustment for them.  Because the remote server won't be able to see data files sitting on their local PC.

 

BASUG is hosting free webinars Next up: Mike Raithel presenting on validating data files on Wednesday July 17. Register now at the Boston Area SAS Users Group event page: https://www.basug.org/events.
merrittr
Obsidian | Level 7

no metadata connections are defined they just start a script and library the sas8bdat provided to them 

Sajid01
Meteorite | Level 14

Input on others moving SAS devs from Base SAS to SAS VA in the cloud
Well is it SAS VA on SAS Viya or SAS VA on SAS 9.4 ?
The answers depend on this question. The architecture of these two environments are substantially different.

merrittr
Obsidian | Level 7
we currently use SAS Base 9.4 with flat sas datafiles produced for the Epidemiologist

we are moving to

SAS VA on Viya (cloud)



so to transition the data files that the epis have all have to be ingested into the metadata server?
SASKiwi
PROC Star

Viya doesn't have a metadata server. The best way to evaluate what needs to be migrated is to download, install and run the SAS Content Assessment Tool.

 

Also you will find the SAS Migration Focus Area useful.

 

SAS Viya is a very different architecture so you would be best to get expert help to migrate if you haven't used it before.

 

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 

Get Started with SAS Information Catalog in SAS Viya

SAS technical trainer Erin Winters shows you how to explore assets, create new data discovery agents, schedule data discovery agents, and much more.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 25 replies
  • 1466 views
  • 4 likes
  • 7 in conversation