Why Data Warehouse Automation kicks the crap out of SSIS

Here’s a great explanation of how to do incremental loads in SSIS. It uses third-party components from Pragmatic Works, and a combination of hand coding and SSIS tasks.

http://blog.pragmaticworks.com/incrementally-loading-data-from-salesforce.com

See how complex it is? How long it takes?

Here’s how to do the same thing in Ajilius:

Simply check the box on the column/s that govern incremental change.

Under the covers we do almost exactly the same tasks as described in the blog post, but the difference is that we do them, not you. You make a decision in the UI, and we instantly generate all the code and logic that implements that decision.

That’s it … a perfect example of why Data Warehouse Automation kicks the crap out of  great ETL tools like SSIS, and why we deliver data warehouses in a fraction of the time spent in those products.

Ajilius. Data warehouses. Faster.

Release 2.4.1

We’re back in the normal release cycle, with our first update to 2.4.

Here’s what you’ll find in this release:

  • New batch scheduler command line parameter “-x” to extract one or more tables to CSV.
  • Printing of script messages in batch output.
  • Licensing revisions making it easier and faster for us to create evaluation license keys.
  • Support for delimited files with spaces in file name.
  • Error corrections.

Internally, we added six new automated test cases this week, around delimited file names, and SCD2 dimension processing. Our automated testing process is really starting to pay off, none of the errors we fixed this week were reported by users, all were found through automated testing.

Registered users will receive an email in the next 24 hours with the download details.

 

Ajilius 2.4.0

We’re a month late, and we’re sorry. But Ajilius 2.4.0 has finally been delivered.

We’ll blog more about specific features, but here’s what you have to look forward to in this release:

  • BI Accelerators for Yellowfin, Tableau and Qlik. Full business-friendly metadata generation at the click (Qlik?) of a mouse. We’re the only data warehouse automation platform to support all three platforms.
  • BI Views for a unified query experience across any BI tool, including PowerBI and Excel. Now your PowerPivot users can share the same business-friendly table and column names as their Yellowfin and Tableau colleagues.
  • A revised, high-performance CTAS engine for MPP platforms, including Microsoft APS/PDW, Azure SQL Data Warehouse, Redshift and Snowflake. And we’re still the only data warehouse automation platform to support three click migration between supported platforms.
  • Full support for Microsoft APS/PDW. That makes us the only data warehouse automation platform that supports EVERY Microsoft RDBMS, SMP and MPP, on-premise and cloud.
  • A new Inference engine for managing early arriving facts. You have the one-click choice of automatically inferring dimension rows when new values are found in fact processing, or simply assigning those rows to the “Unknown” value.
  • Integrated Authentication for all Microsoft sources and targets.

Now the work begins on the Version 3.0 series of releases!

 

Why SQL Server on Linux Rules (again)

I just upgraded three servers. It took less than 5 minutes.

On each server:

sudo apt-get update
sudo apt-get install mssql-server

No more gigabyte downloads (it needed 145Mb), no bewildering and befuddling multi-page installer screens, just two simple commands and the job was done.

On Linux, at least, SQL Server has gone from one of the most complicated installers in the industry to one of the most elegant.

I love it 🙂

 

Yellowfin Integration

One of the challenges in self-service BI is that you can quickly end up in your BI tool’s version of Excel Hell. That is, every user has their own definition of data, sources and aggregations, leading to the pre-DW situation where no two reports ever agreed.

Ajilius now brings a new level of governance to self-service, by directly integrating the data warehouse metadata with the BI tool. Over the 2.4 development cycle we’re bringing integration to a number of popular BI platforms, beginning today with Yellowfin.

yellowfin1

Select a fact table, select a destination, and a full Yellowfin model for your star schema will be written to that destination.

Use the Import function in Yellowfin, and your model is imported as a Yellowfin connection and view.

Not only do we generate the metadata, but Yellowfin makes round-trip integration a breeze. In our demonstrations we show how models can be enriched in line with changes to the data warehouse, preserving your existing reports and dashboards.

Now your data warehouse and your BI solution can be closer than ever before.

Ajilius. Self-service and governance.

DW DBMS Changes

DW DatabasesIn line with DW market developments, we have made some changes to the databases supported by Ajilius as target data warehouse platforms.

The revised list of targets is:

  • SQL Server (On-Premise & IAAS)
  • PostgreSQL / EnterpriseDB
  • MariaDB Column Store
  • Exasol
  • Azure SQL
  • Azure SQL Data Warehouse
  • Snowflake Elastic Data Warehouse
  • AWS Redshift

With the release of SQL Server 2016 SP1, and the CTP of SQL Server on Linux, we’re excited by the potential for major cost savings in mid-market customers. Features like database compression and column store make it feasible to run a small data warehouse / data mart on even SQL Server Express Edition; stepping up through Standard and Enterprise Editions as storage and performance expectations grow.

We have been working with Microsoft on SQL Server on Linux for some time now, and see a bright future for this platform. Installation is easy, performance is great (although very slightly behind Windows at this time), and it has been rock-solid for over 2,000 cycles of DW deployments using Ajilius.

We’ve also been impressed by MariaDB’s new Column Store engine. We’re seeing great performance and scale from this platform, making it a natural target for any business running MariaDB / MySQL as its standard DBMS.

Exasol recently announced V6 of its in memory DBMS, and we’re busy upgrading our Exasol adapter to take advantage of its new features.

With all these new and upgraded platforms comes a harder decision, to drop one of our supported databases. While we introduced Greenplum support earlier this year, we have had just one enquiry during that time, and no customers. Time to retire it, and put the support time to better use.

We’ve still got one platform that we want to add as a target, but I’ll save that announcement until development is further advanced. A hint, it will be a BIG announcement.

Ajilius. Supporting tomorrow’s DW platforms, today.

#1 in Data Sources

Ajilius, the world’s most innovative provider of Data Warehouse Automation, is proud to announce that we are now partners with CData, the world’s leading provider of data access and connectivity solutions.

This partnership brings Ajilius direct data sourcing from applications including Salesforce, Dynamics, SAP, Quickbooks, Xero, Reckon, Marketo, ServiceNow; social media integration with Facebook, Twitter and LinkedIn; and information sources such as OData, OFX, PayPal and eTrade.

A full list of the sources now fully supported by Ajilius may be found at http://www.cdata.com/jdbc/

So now, access to the world’s most popular business data processing systems and information sources is just a few mouse clicks away from your data warehouse.

We’ll progressively roll out support for these drivers over the next 1-2 months, please contact us if you have an urgent need to prioritise a data source.

Ajilius. Unbeatable connectivity in Data Warehouse Automation.

Licensing Change

As our customers grow in number and size, and as the features included in Ajilius continue to grow, it has become necessary to make some changes to our licence structure.

The growth in our customer base has increased our support costs, particularly as we have added new data warehouse platforms.

Our new licence fee increases from USD10,000 per year to USD15,000 per year, from January 1, 2017.

No existing customer will be affected by this change. Ajilius customers have a guarantee that the terms under which they licence Ajilius will be maintained for the life of their usage of the product.

No outstanding proposals will be affected by this change. Any customer engaged with Ajilius before January 2017 will have their existing proposal applied to purchases after that date.

No partner terms and conditions will be affected by this change.

Even though our prices will increase, the value of Ajilius continues to grow. We still maintain our commitment, that you will fully repay your investment in Ajilius on your first data warehouse project.

Ajilius. ROI in weeks, not years.

 

SharePoint Data Warehouse

Reference data – the common lookup tables that drive many data warehouses – often has no clear home in an organisation. It usually gets dumped in spreadsheets, XML files or text files, and ends up in a mess. Recently we’ve found that SharePoint Online makes a great repository for DW reference data. It can be managed and maintained in a clean, multi-user environment, then neatly integrated into the data warehouse.

This was supposed to be a post about the relative merits of SharePoint and Google Sheets for the maintenance of reference data. Unfortunately, Google’s approach to browser-based OAUTH authentication doesn’t play well on servers, so we’re temporarily shelving Google Sheets as an Ajilius data source until we figure out a workable solution.

On with SharePoint!

Adding a SharePoint Online data source to Ajilius is a breeze. Simply enter your site, user name and password:

sharepoint01

Refresh your SharePoint metadata, and you’ll see the libraries, lists and other data sources that can be loaded into Ajilius. Here we’ve created a simple list in which we can maintain details of public holidays.

sharepoint02

You’ll notice that there are many metadata columns added by SharePoint. In fact, we’re only interested in three columns, and we’ll clean that up once we’ve imported the metadata.

Import your metadata just like any other Ajilius data source, by selecting the Load Metadata option from the context menu for the SharePoint.PublicHolidays table. Here is the screen to complete the metadata:

sharepoint03

When added, you’ll see a screen like this, showing the columns for which metadata has been imported:

sharepoint04

Let’s clean up those extra columns we don’t need. Click the Change link at the top of the right-hand panel, and you’ll be taken to the column list.

sharepoint05

We’re going to use the Delete link, highlighted in the previous picture, to quickly delete a large number of columns from the table. When you click it, you will see a simplified form of the column mapper. You can choose to delete a column by changing the drop-down to Delete, either by selection, or most browsers support simply tabbing into the drop-down and pressing the ‘D’ key to set the value.

sharepoint06

When you’ve deleted the columns you don’t need, you should be left with a Date, Store and Title column. Clean up their metadata with some better descriptions, set the Date and Store as business keys to the table, and it should look like this:

sharepoint07

Next, use the Scripts option from the load_public_holiday context menu, and Create and Load your new table. Here is how your screen should look after loading the data.

sharepoint08

And when we view the data, we can see that it has been successfully loaded from our source in SharePoint Online.

sharepoint09

SharePoint Online makes a great multi-user tool for managing the small reference data sets that often get overlooked in the data warehouse governance process. Ajilius now includes a SharePoint connector as a standard feature, at no additional cost.

Ajilius. Now with SharePoint.

 

 

 

Persistent Staging Tutorial

It almost qualifies as a Frequently Asked Question in demonstrations:

How do you handle persistent staging?

Persistent Staging is typically used to keep data from source systems in its original state.

One common reason for doing this is to have a full set of source system data available to support future changes that you may wish to make to star schemas. Having the data already in the data warehouse makes it simple to recreate or modify the star schema at will.

There are two types of persistent staging supported by Ajilius:

  • Latest
  • Historic

Read the tutorial to discover exactly how they work.

http://172.104.56.212/persistent-staging-tutorial/