DavidHenderson Tracker

Re: Regarding :: which log to check for data load in memory.

DavidHenderson — Fri, 01 Aug 2014 13:07:52 GMT

Unfortunately, there isn't a log where this information is written. The good news is that the new release of SAS Visual Analytics (7.1 - shipping this fall) will include what you are looking for right there in the UI.

In SAS Visual Analytics Administrator 7.1, the size of the table (in MB) is listed on the LASR Tables tab. The SAS LASR Anayltics Server has also added the ability to compress data when putting it into memory and VA Administrator and Visual Data Builder will both support that functionality in 7.1 as well.

There are some other cool features concerning memory management and governance coming in 7.1-- stay tuned for more details as we get closer to the release date.

Re: Importing Data from SAP to SAS Visual Analytics

DavidHenderson — Thu, 31 Jul 2014 15:24:50 GMT

@sat_lr, You can SAS Visual Data Builder to create queries to pull data from any SQL database using SAS/ACCESS to SAP. I know that is a pretty basic answer and having seen other posts and responses from you, I get the feeling that you are looking for more detail. I'd be happy to provide those details if you can give me a better idea of what it is you are looking for or what problems you are trying to solve.

-David.

Re: Issue starting LASR server

DavidHenderson — Mon, 21 Jul 2014 15:44:49 GMT

Please contact tech support about this issue. They will need additional information about your system, including what version of Base SAS and SAS LASR Analytic Server you are running.

Re: SAS BI Dashboard viewing offline

DavidHenderson — Mon, 14 Jul 2014 19:15:42 GMT

SAS BI Dashboard is a web application. As with other web applications, it isn't possible to view it offline.

Re: VA, HDFS and Cloudera

DavidHenderson — Wed, 25 Jun 2014 17:29:26 GMT

No, I am saying that data stored in a typical Hadoop format isn't the format needed by LASR to map it directly into memory. SAS/ACCESS for Hadoop becomes the method through which that data is read from its form in CDH and then put into LASR (either directly into memory using a parallel load of streamed in from the workspace server, or by writing it into LASR's HDFS as a SASHDAT file and then mem-mapped into memory.)

Re: VA, HDFS and Cloudera

DavidHenderson — Tue, 24 Jun 2014 14:49:22 GMT

SAS LASR Analytic Server's use of HDFS deserves a bit of explanation here...

If your data is sitting in your corporate deployment of Hadoop, say Cloudera, and is used for non-LASR activities (such as Pig or Hive jobs), it isn't in a form that can be used directly by LASR. When you use SAS to move data onto the co-located HDFS that is used by LASR, the resulting data file (.SASHDAT) is very specific to what LASR needs. Once in that format, it isn't usable by other Hadoop processes, but it is very efficient to load those files into LASR because it is laid out on disk exactly the same as it is in memory. LASR uses a Linux process called memory mapping to map the locations on disk into RAM memory. The additional benefit is that if the OS needs to page that memory out, it doesn't have to write it to disk-- it's already on disk and the OS knows where. So the OS can just drop it from memory during the page swap. When it comes time to swap back into memory, the OS picks it up from the original location on disk.

If you want to, you can install LASR directly on your corporate deployment of Cloudera (or Apache Hadoop or HortonWorks or BigInsights, etc.). However, even if you do, you'll need to decide if you want to create SASHDAT files in addition to keeping your data in its traditional form for use by Hadoop processes. The main benefits are listed above. The drawbacks are additional disk usage (more than double in many cases) and the need to somehow keep the two data sources in sync.

Re: Physical Storage of Reports

DavidHenderson — Tue, 24 Jun 2014 14:17:55 GMT

, Jaap is correct-- A report is never just stored in a directory on disk. The "file" is stored in the SAS Content Server. How that information is stored in SAS Content Server has changed over time. Initially the content server used a DAV to store the files, but now that content is in a database. Additional information about the report is stored in the metadata server.

Re: Icons on Home page

DavidHenderson — Fri, 23 May 2014 18:59:55 GMT

That would potentially work if all users had the same access to the report. Because of table security and even row-level security, users can potentially see very different data in the same report. As a result, a thumbnail created based on your data and shown to me presents a security risk. That means that using a single thumbnail for all users isn't a valid solution if the thumbnail shows real data. That is why I suggested the use of stylized icons.

Re: Icons on Home page

DavidHenderson — Mon, 19 May 2014 19:35:50 GMT

What you are asking for isn't available at this time. It's a great idea, but one that is a bit challenging to implement. rendering pseudo-realistic thumbnails of each report requires retrieving data and running the same analysis as opening the first section of each of the reports. If even just a few dozen users were hitting the Home page at the same time each with ten or so favorites and recents, the system would have to renders 100's of reports.

Would rendering the first section of the report using stylized graphics suffice? For example, instead of seeing a very generic report icon, you might see the real title with stylized graphs/tables that indicate what type of object it is, but don't reflect the true data.

Re: View multiple reports

DavidHenderson — Mon, 12 May 2014 15:23:08 GMT

There is not really a way to view multiple reports in a single browser tab in the SAS Visual Analytics Report Viewer application. However, if you wanted to get creative and use iframes in your own web page, you could show multiple reports. See my comment on this other thread for details:

https://communities.sas.com/ideas/1506

Hope that helps...

-David,

Re: Mobile BI Offline Capability

DavidHenderson — Thu, 08 May 2014 18:05:13 GMT

@satlr, whether the data is stored on the mobile device (whether iOS or Android) is dictated by the capabilities assigned to your user ID. Based on what you are reporting, I assume that your administrator has it set to not permit you to store data on the device.

Re: VA Performance results/experience with a concurrent workload

DavidHenderson — Thu, 08 May 2014 17:57:53 GMT

I think I was able to find exactly what you are looking for. I posted the document to the forum as an article:

Hope that helps!

SAS VA 6.3 and LASR SMP sizing guidelines.pdf

DavidHenderson — Mon, 05 Oct 2015 20:47:34 GMT

We have often been asked about how many users a given hardware configuration can support. This paper, written by the SAS Enterprise Excellence Center provides clear information on load capacity for a SAS Visual Analytics running with non-distributed SAS LASR Analytic Server. Although the paper discusses VA 6.3, the same information applies to 6.4.

Re: VA Performance results/experience with a concurrent workload

DavidHenderson — Thu, 08 May 2014 14:32:10 GMT

, unfortunately, I can't give you a formula to calculate what kind of load a given number of users will place on the system. There are just too many variables-- not only does number of users, data size and schema, and hardware play a role, but the activities of the users is a huge factor. I can create a report with simple tables and graphs that perform little more than summarizations or I could create a report that is exported from VA Explorer with heat maps and decision trees that creates significant more load on the system. If I then have 20 users just viewing each of those reports, one group puts much bigger load on the system than the other, even though both groups are just "viewing reports".

And although I hope that helps to understand why it's near impossible to provide a formula, it doesn't answer your question.

6 GB isn't much data and 20 users is more than reasonable for that system-- You will be fine. I will pull up some of our performance testing numbers and post them later today.

Hoffentlich, hilft das etwa!

Re: Loading the Reports in SAS VA

DavidHenderson — Tue, 15 Apr 2014 16:20:51 GMT

Again, I understand that you are the only one on the system and as a result, the concerns for which we architected don't necessarily apply, but a single person person working on a demo system isn't exactly our targetted audience.

Re: SAS VA and comments limitations

DavidHenderson — Tue, 15 Apr 2014 16:16:13 GMT

I checked on the limit on attachment file size. As far as I can tell, there is no files size limitation.

Re: Loading the Reports in SAS VA

DavidHenderson — Mon, 07 Apr 2014 17:01:57 GMT

Although I understand your concern about having to wait, I'd also like to point out the other side of the argument...

By pre-loading every section of every report that every user opens, you are assuming the cost of all of that analysis regardless of whether or not the users choose to look at every section of every report they open. That would not be a concern if processing resources were unlimited, but that is the rare case. On a typical system, CPU is a contentious resource, meaning that when one user's job runs, other users inevitably are impacted. Loading all the other sections in your report as you suggest (and doing the same for every other user) would almost definitely have a negative impact on the overall performance of the system.

Re: SAS VA and comments limitations

DavidHenderson — Mon, 07 Apr 2014 16:11:39 GMT

, I can answer most of your questions immediately, but I need to do a little research on one of them...

1. Comments are stored in a separate database, so they are not tied directly to the report. The advantage is that a user doesn't need write permission on the report to add a comment about it. Additionally, unless the user chooses to view the comments about a given report, he/she doesn't incur the overhead of fetching all of those comments.

2. This question is pretty broad, so I am not sure of what you are wondering about. One limitation is that because the comments are stored separately, they aren't included when the report is exported. You also need to be careful when moving a report from folder to folder... if you make that move by copying, then deleting the original, you will probably lose the comments.

3. There is no limit to the number of comments that can be added to a report.

4. Comments don't have expiration dates, so they remain until someone does something to delete them.

5. I don't know what the file size limit is... I'll check on this and get back to you.

6. I don't have any other recommendations... It would be helpful if you explain what you are attempting to accomplish.

Re: Reports and Permissions

DavidHenderson — Tue, 25 Mar 2014 16:09:16 GMT

@mjaitly, statement is correct, although don't forget that single table level security may provide what you want. If you deny a given user access to the underlying data tables, objects that use that table won't display their data. It isn't quite as clean as what you are asking about, but it provides similar functionality.

There are no plans to implement section-level or object-level security at this time.

Re: Suggestions for demonstration hardware (SMP) for Visual Analytics

DavidHenderson — Tue, 18 Mar 2014 17:25:06 GMT

If you are really only interested in demoing VA and can limit your data, you can get away with running the entire VA stack along with the LASR SMP server on a beefy laptop with 4+ cores and 16 GB of RAM. This would not be an environment that would support multiple users or large data, but that should be adequate for demos. Note too, that such an environment would not be supported by Tech Support. If you run into problems that are possibly caused by the small env, you would have to replicate them on a supported configuration.