Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have a project with a 2TB database (450,000,000 rows). What is the recommend memory? There will be about 200 users at the same time using the info.

share|improve this question
    
450M rows = 2TB = ~ 4.7kb per row, including indexes. Exclude indexes and that's a pretty small row size. With 200 concurrent users, that's going to be a fair amount of data in cache... –  Mark Henderson Aug 27 '09 at 22:20
add comment

3 Answers

up vote 7 down vote accepted

You can go to 32 GB on 32 bit (assuming Enterprise), far more in 64 bit.

As much as you can afford, basically, within the SQL Server edition limits.

The data size itself is neither here not there: it's how much is in use at any one time

share|improve this answer
1  
RAM is cheap. You'll most likely be using FB-DIMMs in a database server with that much data, but even so, 16Gb kits are a few hundred dollars. There's no way you want to be doing that in 32-bit, as you'll most likely have 64-bit row identifiers, so just put in as many of the largest dimm's you can find! –  Mark Henderson Aug 27 '09 at 22:22
add comment

Q: "What is the recommend memeory?"

A: "The Most"[1]

[1] For server hardware where read access is a top priority and the data is highly cachable.

share|improve this answer
add comment
  1. How much can you afford?
  2. Can you squeeze for more money and get more than that?
  3. It's still not enough. ;-)

More RAM means more data which can be kept in memory. More data in memory means less data to fetch from disk. Less disk activity means better performance.

The usage patterns of the application/s accessing the database are more important than the size of the DB, number of rows and number of users.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.