Hardware requirements?

Hey all,

I couldn't find an answer anywhere on the site, if there is one, could you please let me know where to go? Thanks.

Here goes. We receive large csv files, around a million records a month. What we need to do is:
- Of course, import these records
- Be able to run failry heavy reports on around 3 months worth of data (about 3 million records)
- Wanting to store at least a years worth of data

What I need to know:
- What server(s) spec (CPU, HDD, memory, etc) would we need to import and run reports quite fast. Don't want the reports/imports to take days to run. The reports will need to be run by about 3-5 people at the same time
- What OS and database type should I use (I was looking at Windows/Linux and MS SQL/MySQL)
- If we run these types of reports, can we use the same SQL server to run/store other applications (eg. CRM, etc) with no real impact on performance. There would be at least 20 people using the CRM database at the same time.

ThanksHowdy

You need a serious server - HP DL760 quad CPU 16GB RAM ( min ) 2 TB of storage ( or bigger) .

Windows 2003 and SQL 2000 Advanved Server. $$$$$$$

Any other OS's and servers will do same job,. Depends if you want windows or freeware...

Cheers,

SG|||Thanks for that SG. That is some serious power needed.

Just to understand how long it will take. I have a Compaq DL380 with 2GB RAM. Can you tell me more or less how long it will take to generate a report grouping a specific field and doing a total off another field, assuming it is based on the 3 million records?

Thanks|||Huh?

I don't claim to be a hardware expert, but SQL Server shouldn't even sneeze at 3 million records on a mid-range box.

2 TB storage? For 12 million records a year? Assuming about 500 bytes per record that is 180 years of storage.

16 GB or RAM seems like a steep requirement too, but RAM is never wasted on a server...

As far as how long it takes, my desktop SQL Server will churn through 3 million records in less than a second.

Any one else with opinions on this?|||SQL Server Express...

http://blogs.msdn.com/sqlexpress