January 9, 2011 at 8:17 am
I was just reading Tony Davis' editorial on RAID 5 and SQL Server and it's beginning to dawn on me that I'm beating myself up trying to learn storage and the intricacies of things like plan cache, index padding, statistics maintenance, etc etc when what interests me is loading sql server from wildly disparate data sources and extracting data into reports useful to analysts and decision-makers.
Related to this ( I think ) is the trend towards dividing up traditional "DBA-duties" amongst several groups or individuals. It reminds me of the specialties we see in the medical field. Can you really be top-notch at storage, database design, Tsql, reporting, data warehousing etc?
Besides the trend towards "no traditional DBA in-house," at my company and others, I also see a situation where the database staff ( DB developers, architect, acting DBA (me) ) has little if any access to things like storage and often little input to the decisions about things like storage and other hardware. This may simply be a dysfunctional situation in our case where turf wars lead to Systems types wanting total control of anything viewed as "hardware-oriented" with no DB staff interference.
In any event, it may be time for me to shift focus towards things like warehousing and reporting services.
January 9, 2011 at 9:25 am
I wouldn't necessarily jump off a building yet. There is specialization, and this product is so absolutely huge, that no one is going to know everything. But you can be a generalist have a great career. You don't have to specialize. That said, specialization brings benefits too. And thinking about focusing on SSIS & SSRS is probably a very good idea. What I've seen over the last 9 years at the large company where I currently work is that we've focused mostly on building mechanisms for collecting data for most of that time. Now, they actually want to consume the data collected and start makin decisions from it. The growth, at that company, is in BI. But I'm aware of other companies that are just starting the rebuild their data collection. It's a cycle, it all comes around.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
January 9, 2011 at 1:40 pm
I think we would be safe in making some generalizations. Many companies using sql server just aren't going to budget for a full-time DBA until they reach the size and complexity where such person can actually keep busy full-time on DBA tasks. For these ( small to medium size ) companies it may be the consultant who rides to the rescue on occasion as needed. ( if management is smart enough to hire the consultant AND follow their advice ).
One of the problems is when systems are designed without the benefit of anyone on hand who really understands what will be needed in terms of design and hardware when the system reaches full-size ( e.g. after all data is migrated from legacy systems ).
When you've been operating with a few hundred foxpro databases ( one per client ) spread over ten servers running IIS I think you can be lulled into a false notion of what's going to be needed when all of that data is in one database on one system ( this is our case ).
I could go on, but I think the bottom line is having a CIO/CTO on board from the beginning who REALLY understands sql server and has deep experience scaling up a system. Without that, the managers making decisions can be snookered by existing Systems Admins and database developers who really don't have that experience.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply