About LinusH

LinusH · ‎02-03-2026

When you say storage, do you mean SW and data on disk? Disk is usually cheap (unless you have humongous sized data) and usually easy to augment over time. Cores and RAM usually drives cost. Apart from contacting SAS, here are some useful (?) links: System Requirements: https://documentation.sas.com/doc/en/itopscdc/v_072/itopssr/titlepage.htm?fromDefault=#page=9 Sizing Recommendations: https://documentation.sas.com/doc/en/sasadmincdc/v_072/itopssr/n08i2gqb3vflqxn0zcydkgcood20.htm Sizing Prez: https://communities.sas.com/t5/SAS-Explore-Presentations/Sizing-your-SAS-Viya-4-Environment/ta-p/838247

LinusH · ‎02-02-2026

I mean, why don't you just try?

LinusH · ‎01-30-2026

Not specifically, no. Again, maybe engage with you local SAS team or partner. If you have a specific question/use case you can try to formulate an inquiry here on communities, probably under the Solutions/Data Management.

LinusH · ‎01-30-2026

I have a client who has multiple SAS Web App Servers. On of them has products currently not in use. And we wish to save resources by not starting it. If I understand correctly the sas.servers.mid script uses a simple /bin/echo SASServer* It seems easy just to rename the SASServerX_1 directory. But does that any unwanted side effects? I guess the alternative is to unregister those products using the SAS Deployment Manager, but that seems a bit...final... We are on RHEL, 9.4M7 and M8. Any thoughts/recommendations are appreciated.

LinusH · ‎01-28-2026

@domroy, I can't find any clear migration path either. I seem that Federation Server is supported in Content Assessment, so if you haven't run that already, do. And I would contact your SAS representative with the overall strategy and plans for Federation Server sites.

LinusH · ‎01-27-2026

Column length up 32 is no problem for SAS, but column name length has the same limitation as table name. If you have tables that you know comply with SAS naming standard (also beware of special character, spaces, initial digit etc) and they're not visible to SAS - try to connect to the database using the same credentials from another client using the same connection configuration.

LinusH · ‎01-16-2026

This step uses SQL, and SQL does not aggregate based on formats. So what you need to do is transform this column, by using put(LaDate,YYWEEKV5 .).

LinusH · ‎01-08-2026

I don't have access to Viya 4 at this point. But it seems there is no OOTB functionality for this. It seems that you could implement this on a Flow instance level, by pointing out a program to execute when a job flow exits: https://documentation.sas.com/doc/en/sasadmincdc/v_071/evfun/n0b9cf8ru47gp6n1lvamxqwbr3by.htm#p00388x4a7b6ywn1ce2d03tz88ns

LinusH · ‎01-08-2026

Ok, I don't have access to Studio Engineer at this point to try myself. What does the "i" symbol say if you click/hoover over it? You should compare data type and columns lengths for your datasets.

LinusH · ‎01-08-2026

It's hard from just one screenshot to tell what's wrong. Maybe you could attach the generated SAS code?

LinusH · ‎12-19-2025

Placing office files in SharePoint and in Teams let you edit the file simultaneously, perhaps this also works for SAS access? Haven't tried since my SAS server can't access those due to firewall rules.

LinusH · ‎12-19-2025

SAS is communicating with Git from the DI Studio client, not the server., and this is where the local repo is kept. What does the git config --global -l command give you? Did the "initialize Repository" work without any messages?

LinusH · ‎12-15-2025

Haven't used much. But a pro for NFS that you might need as much total disk for your work. But if performance is important, chances are that you probably can get faster disk access with locally mounted disks. To verify, test and/or discus with your infrastructure team.

LinusH · ‎12-09-2025

You might find some formulas that calculates this. But the easiest way is to take a subset and test yourself. Things might affect result if you have compression on your data and indexing. I'm usually pro long dataset over wide, but this is an extreme, key-value-pair model which is usually not useful for human usage. What is the purpose of this re-modelling?

LinusH · ‎12-08-2025

You macro variable is over 3100, so $2500 is not enough. I see that you also have double spaces, added compbl() if you want to get rid of those. data _null_; length b_allv_Y $4000; b_allv_Y = "'" || tranwrd(compbl("&allv_Y"), " ", "' '") || "'"; call symputx('b_allv_Y', b_allv_Y); run;