Architecting, installing and maintaining your SAS environment

copying datasets between servers (dev, test, prod)

Accepted Solution Solved
Reply
Occasional Contributor
Posts: 13
Accepted Solution

copying datasets between servers (dev, test, prod)

I would really appreciate some help in moving dataset from PROD to either DEV or TEST.

 

Each environment has its own Lev on seperate metadata ports 

 

DEV - 8563 - LEV3

TEST  -8562 - LEV2

PROD - 8561 - LEV1

 

We have a large PROD area which houses about 25TB, for DEV and TEST there is only 3TB. So obviously we cant keep all datasets in all three environments.

 

In our development BAU, a developer may want to copy a dataset from PROD to DEV, maybe a exsisting mart to query from.

 

Now the easy way would be to get them to winscp or x command the data over, but i want to use SAS functions for security (metabound), contraints like indexes. in old worl when we were on one server, i could just proc datasets or proc copy, but now we are different servers (in metadata's eyes) its not that simple.

 

Forgive me but i never had to use older versions of BASE SAS to connect to remote servers etc, i know there are functions like PROC DOWNLOAD, RSUBMIT. so i was hoping someone help me so i could write a stored process so a user could select libs and associated tables.


Accepted Solutions
Solution
‎02-20-2017 04:42 AM
Super User
Posts: 3,254

Re: copying datasets between servers (dev, test, prod)

Posted in reply to sasadmin_tlove

If it's true all your servers are the same OS and SAS version, you don't need SAS/CONNECT and UPLOAD and DOWNLOAD. It would be a lot simpler to use PROC DATASETS where you can copy a whole library in one COPY statement via the use of network shares.

View solution in original post


All Replies
Super User
Posts: 3,254

Re: copying datasets between servers (dev, test, prod)

Posted in reply to sasadmin_tlove

If you set up network shares between your servers (I'm assuming here they all have the same OS and SAS version) then you can use either OS copy commands or SAS procedures like PROC DATASETS.

SAS Super FREQ
Posts: 299

Re: copying datasets between servers (dev, test, prod)

Posted in reply to sasadmin_tlove

HI,

 

In addition to what SASKiwi suggested, an additional thought:

 

Your users copy their own data across the environment, correct? Does this happen quite frequently, or is it

rather an update "mechanism" at the end of a day?

Reason I am asking is to see whether promotion would be an option for you?

(export / import wizards).

 

The users can save their work in their dedicated SAS folders (granted you created all content in metadata),

and at the end of the day you could run an expert / import across environment.

You could automize it via batch tools.

 

Thanks

Anja

Occasional Contributor
Posts: 13

Re: copying datasets between servers (dev, test, prod)

Hello,
Thanks for reading.
At this point, I suppose I just want to understand, if I'm logged into PROD, how do I use proc download to copy a data set to the DEV location, in the same library. From an O/S point of view, the dataset and indexes would be copied to the desired DEV Library location. From a SAS Point of view, the Librarys have the same name, the library locations are different, and the 2 environments sit on 2 separate metadata port.
The stored process I'm building, uses dynamic prompts to tell the users what tables and lib are available in prod, but then I'm not sure of the syntax, either rsubmit or proc download to copy the datasets to the target library,
Solution
‎02-20-2017 04:42 AM
Super User
Posts: 3,254

Re: copying datasets between servers (dev, test, prod)

Posted in reply to sasadmin_tlove

If it's true all your servers are the same OS and SAS version, you don't need SAS/CONNECT and UPLOAD and DOWNLOAD. It would be a lot simpler to use PROC DATASETS where you can copy a whole library in one COPY statement via the use of network shares.

Occasional Contributor
Posts: 13

Re: copying datasets between servers (dev, test, prod)

This makes a lot of sense. Sometimes when you look too closely at something, you can miss the obvious.

 

 

☑ This topic is solved.

Need further help from the community? Please ask a new question.

Discussion stats
  • 5 replies
  • 429 views
  • 3 likes
  • 3 in conversation