Archive for month: June, 2013
The title should actually read “8 Random Ways to Optimise SSIS”.
One of the recent project I have been working on involved building a distributed (scaled-out) SSIS environment, this means multiple VMs with a standalone SSIS (2012 in Package mode) instances installed (so no SQL Server Database Engine), all pushing massive amount of data to a staging database.
I have been brought in on this project to suggest a few techniques to improve the performance of the scaled-out SSIS environment, by basically increasing the throughput to the staging database, below I discuss some of the general approaches I have taken to achieve that goal. Some of the advice might be a bit random, and others might not be pertinent to your particular situation, but over-all you should find a gem or two on optimising SSIS performance in there somewhere!
Its a RDBMs, no its a ColumnStore, no its a cloud integrated storage platform… Oh its actually SQL Server 2014!
Exciting news as the new (project code-name: Hekaton) SQL Server 2014 is community technology preview (CTP1) is out and can be downloaded and evaluated by the community.
This is a pretty simple post to show how to alter (add, remove or edit) a calculated field in an SSAS cube without redeploying the whole project, a useful technique if you do not have the SSAS cube project handy or wish to quickly implement changes on a live cube.
Until now, I have found working with Kerberos when setting up a SQL Server stack to be a complete nightmarish experience, mainly due to two reasons:
- Working with Kerberos usually requires access rights to Active Directory for the account setting up this authentication protocol on the stack, in order to be able to effectively diagnose the setup and also configure the Service Principal Names (SPN) for the various SQL Server and SharePoint service accounts, and setup delegation. This means SQL Server architects and Network Administrators need to collaborate in order to correctly configure the stack, which is often an unpleasant and long winded experience of trial and error.
- The lack of a centralized diagnostic and configuration tools for Kerberos setup on SQL Server makes this tasks very tedious, particularly if you follow the limited number of online resources out there to setup Kerberos, and find that they do not apply exactly to your situation, or do not work exactly as intended after following the lengthy steps, and you are left with a very limited option in terms of diagnosing exactly what went wrong.