Since this option doesn't work in the SAS 9.4 Snowflake connector I instead use this to trim VARCHAR columns to a reasonable size:   conopts = 'default_varchar_size=256' default_varchar_size is a Snowflake option you can set from SAS.   Also I have the COMPRESS SAS option on to ensure Snowflake tables read into SAS don't blow out in size.   ... View more

Apologies, I meant DISTINCT. No doubt you've realised that now.   ... View more

Typically your Option 2 Passthru should be quicker as you can be sure filtering happens in Teradata before downloading. However why are you using DISCRETE which will only slow processing? I never use this if I can avoid it.   Also how many rows do you expect from this query? Even if your query selects only 1% of the rows that is still 8 million and will take significant time to select and download to SAS. Do a COUNT(*) first to see how many rows will be downloaded.  ... View more

That sounds good for 5 million rows. ... View more

Add INSERTBUFF = 10000 to your LIBNAME statement - does that speed things up? Also you need to explain what you mean by duplicate records. Does that mean entire rows are duplicated or are you referring to rows where only the primary key is duplicated? I would imagine defining a unique primary key would only allow the first row where the key is repeated to be inserted. ... View more

First step is to connect SMC to your SAS metadata server by creating a connection profile like this:   ... View more

Are you using a DATA step as well? What's your INSERTBUFF setting if any on your Redshift connection?   ... View more

You are using SAS SQL so the correct way to define a date constant in SAS SQL is: '15Jul2025'd   ... View more

BTW in my experience running a different version of the Windows SAS client tools compared to the server software doesn't cause significant problems and everything should still just work. It is still good practice to align versions, but for example where I work we run multiple versions of SAS server software, but only use the latest version Windows SAS client tools with the multiple versions. ... View more

If you are OK with just email notifications from a scheduler running your programs then no addtional coding is required and you only need to setup the notifications similar to this:         ... View more

@HemanthMG - Looks like you are on the right track. Another thing to watch out for is encoding differences. Most new SAS installations now set the default encoding to UTF-8. If your source installation isn't UTF-8 then you may need to convert your SAS datasets' encoding when migrating. That means you can't use OS tools for moving SAS data.  ... View more

You don't mention your source and target OSs. They will likely have a greater influence on your migration strategy than the underlying cloud platforms. Also will your M8 platform have the same architecture as your M6 one? Have you run the SAS Content Assessment tool on your M6 installation yet? I would regard that as the best first step.  ... View more

I'd suggest putting your concerns in a Tech Support track if you haven't already. You'll gain better visibility that way. Please post their response back into this thread. ... View more