SASKiwi Tracker

Re: SAS VA 9.4 M8 > M9 update fail

SASKiwi — Wed, 20 May 2026 20:03:41 GMT

For this type of problem when there is no obvious evidence of the cause, then a good starting point for a Windows SAS installation would be to check if all SAS services are running on all servers in the installation. If any are not running, then trying a manual start would be the next step. If you are a SAS administrator then these steps should be familiar to you. If not then talk to your SAS administrator to find out what troubleshooting has been done.

I would also follow @JOL 's recommendation.

Re: SAS to Snowflake data load using Key Pair

SASKiwi — Tue, 12 May 2026 21:29:00 GMT

If you use a DATA step with Snowflake then the database, and schema must be defined in the LIBNAME statement. You can only use a LIBREF and a table name in the DATA statement like in the example I posted.

Re: SAS to Snowflake data load using Key Pair

SASKiwi — Tue, 12 May 2026 20:15:48 GMT

PERM is the only LIBREF you have defined for Snowflake. If you want to read a SAS table and write it to Snowflake in a DATA step the syntax should look like this:

data perm.cars;
    set sashelp.cars;
run;

Re: Issues with SAS VIYA caslib to Amazon Redshift -

SASKiwi — Wed, 06 May 2026 20:14:27 GMT

Does the Redshift connection work OK in a DATA step or PROC SQL query? If not then I would suspect the problem is with the Redshift connection. If yes then I would suspect there is an issue on the CAS side.

Re: Reading Unicode from Snowflake into SAS via ODBC

SASKiwi — Mon, 27 Apr 2026 23:59:06 GMT

@AndrewZ - This is indeed the Snowflake driver used for the SAS/Access Interface to Snowflake. It should have posted this link: https://tshf.sas.com/techsup/download/hotfix/HF2/M3H.html#M3H004

Re: Reading Unicode from Snowflake into SAS via ODBC

SASKiwi — Mon, 27 Apr 2026 21:33:09 GMT

@AndrewZ - SAS have just released a new version of the SAS Snowflake ODBC driver for which there is a hot fix available. The previous SAS-supplied driver is no longer supported by Snowflake so it might be worthwhile testing this fix.

Re: Bulk loading

SASKiwi — Thu, 16 Apr 2026 20:09:22 GMT

You should also try setting the INSERTBUFF option as the default value is not optimal. Try INSERTBUFF = 10000 as a starting point but feel free to experiment with higher or lower values.

Re: Is there a way to determine of a SAS programming running (in EG) or has crashed or?

SASKiwi — Sun, 12 Apr 2026 20:25:28 GMT

Please note that when using a remote SAS server it can be quite normal to have large variations in run time with the same program and data. This is because you are sharing the server with other users and if it is heavily loaded then program run times can take a lot longer than usual. The way you are monitoring the size of your lck SAS files is a good one.

Re: How can SAS Web Application users be tracked on a SAS Linux platform?

SASKiwi — Wed, 08 Apr 2026 20:18:03 GMT

A starting point to tracking user activity in SAS installations is to implement SAS's APM functionality. Here is a useful link for this: https://support.sas.com/kb/68/709.html

You should also search the SAS documentation for APM to learn more.

Re: SAS/ACCESS SNOWFLAKE LIBNAME

SASKiwi — Mon, 30 Mar 2026 21:58:00 GMT

Does this help? - https://community.snowflake.com/s/article/ODBC-connectivity-to-snowflake-fails-with-error-Marshaling-private-key-failed-when-using-Key-Pair-Authentication

Re: SAS/ACCESS SNOWFLAKE LIBNAME

SASKiwi — Sat, 28 Mar 2026 21:13:23 GMT

Is your SAS Viya environment already configured and working with Snowflake? If it is then tweaking it for a different database and/or authentication method shouldn't be too difficult. If it isn't then authentication is just one of many things that will need to be configured, including Viya.

What authentication method you use will most likely be dictated by what your Snowflake administrators have allowed. I suggest you consult with them if you haven't already.

Re: Code to Download a File from SAS Studio (Viya)

SASKiwi — Tue, 24 Mar 2026 19:10:43 GMT

As @Kurt_Bremser has already stated, just a script wont work unless your administrators set up a network folder share that is accessible from your SAS server. Once that is done SAS code can be run on Viya 4 that will write to the share. PROC EXPORT would be one option to write SAS data to an Excel spreadsheet or a CSV file.

Re: SAS VIYA - where to start?

SASKiwi — Mon, 23 Mar 2026 23:57:54 GMT

I suggest you explore creating a Viya Visual Analytics dashboard to track metrics from your monthly data feeds. Seeing your data graphically often shows up issues that otherwise would not be obvious. You can start to get ideas for a dashboard from the Viya Visual Analytics Gallery. It doesn't have to be as fancy as many of these. Start simple and go from there.

Re: Authorization errors post Integrated Windows Authentication

SASKiwi — Sun, 22 Mar 2026 05:26:10 GMT

What SAS user interface is used here? If it is Enterprise Guide, then change your EG server connection to not use IWA (just Windows account and password) and test the same LIBNAME again. If this works then the problem is with IWA and SPNs for the storage being accessed. If it doesn't work then the problem is with the Windows account permissions on the folders in the LIBNAME and has nothing to do with IWA.

Re: SAS Enterprise Guide Error Message

SASKiwi — Fri, 20 Mar 2026 22:43:29 GMT

Looks like you are reading all columns from your Snowflake table CLD95_TRANSACTION_MM. How many columns are there? If suggest you try just extracting one column from the table as a test. Do you get the same error? If not try adding columns one at a time until you identify the column giving the error.

Re: Exploring SAS Programming Interfaces Q&A, Slides, and On-Demand Recording

SASKiwi — Thu, 19 Mar 2026 00:23:12 GMT

@LukeDalton - We use full folder paths for our Git repository directories by creating a macro variable that points to the required repo's root directory, then use that macro variable in all subfolder path definitions. Changing to a different repo is as easy as changing the root directory macro variable.

We find that works reliably and transparently for us as long as you are disciplined about using that macro variable. The problem with relative paths is where is the starting directory? If could be a user directory or it could be the folder where your SAS executable is depending on how your SAS installation is configured.

Re: SAS EG needs a function to display the log AND automatically save it

SASKiwi — Thu, 12 Mar 2026 19:07:28 GMT

@Wolverine - I get the impression you are using EG for both running Production jobs where you want to keep SAS logs and also for program development and analytics. If this is the case then an alternative approach could be to move the running of your Production jobs to SAS Management Console. Not only can they be automated with Schedule Manager, SAS logs are produced and saved automatically. Not only that you can trigger automatic emails if the jobs fail so you only need to look at SAS logs if you get an email. We use this approach and it works a treat.

Re: Migration- SAS 9.4 to SAS Viya - Spk file export

SASKiwi — Sat, 07 Mar 2026 02:27:56 GMT

@Somu - Thanks for sharing more details of your migration practices. Of course there are many ways of dealing with this depending on your business requirements and IT infrastructure.

In our case it is essential to be able to run the same applications in both our source and target installations in parallel and to check that we get the same results, right down to key datasets being identical or not materially different. That's why we have near real time data transfers and code synchronization.

Re: Efficiency Matters: Compressing SAS Datasets

SASKiwi — Thu, 26 Feb 2026 20:17:36 GMT

@touwen_k - Run some tests on your typical datasets using both CHAR and BINARY and compare the compression percentage. Go with the one that provides the highest compression. Compression works in Viya.

Re: Importing csv file results in bizarre results in character string.

SASKiwi — Mon, 23 Feb 2026 23:55:10 GMT

Did you save it as a UTF-8 CSV to begin with?

Otherwise using the ENCODING option is unlikely to work.