Subscribe to:
Post Comments (Atom)
Labels
About Me
Best Practices
Career
Data Mining
Documentation
Feature Requests
Humor
MagicPASS
Meme Monday
Mirroring
Parameter Sniffing
PASS
Performance
PowerShell
Presentations
Query Tuning
Recognition
Replication
Scripts
Security
SQL Power Doc
SQL Server 2005
SQL Server 2008
SQLH2
SQLRally
SQLSaturday
SYDI
T-SQL Tuesday
Tips
Troubleshooting
Updates
VirtualBox
Windows
XML
Copyright © 2015 Kendal Van Dyke. All rights reserved.
Kendal is a database strategist, community advocate, public speaker, and blogger. A practiced IT professional with over 15 years of SQL Server experience, Kendal excels at disaster recovery, high availability planning/implementation, & debugging/troubleshooting mission critical SQL Server environments. Kendal is a Senior Consultant on the Microsoft Premier Developer Support team and President of MagicPASS, the Orlando, FL based chapter of PASS. Before joining Microsoft, Kendal was a SQL Server/Data Platform MVP from 2011-2016.
[About Kendal]
(http://lh3.googleusercontent.com/-wrW1fk8IiFE/Vr36w9dtRxI/AAAAAAAADCw/tVa4vTgLWIw/w139-h140-p/IMG_3503.JPG)
5 comments
Nice review. Thank you.
Excellent, is very interesting. Good job.
Very interesting info - and I assume its also fairly accurate. BUT I don't see the point of the PHYSICAL test that you performed. If the point to compare a single RAID 10 disk to 2 RAID 1 disks, you have to find a way to do just that.
Testing a single RAID 1 disk and doubling the performance (or using a data file that is half the size) doesn't prove anything.
Again - I think you conclusions are correct, but I don't see the point in running the physical test when you are "calculating" your results anyway.
Wow, great benchmarks you got there. I would like to know, if you had 6 disks, setup as 3 RAID1 sets (2 disks each) to divide data, log, and tempdb, would you choose that one over a RAID 10?
To previous anonymous post -
The one factor that would need to be compared is the concurrency of reads/writes. If you're going to have the data/log/tempdb on a single set of synchronized spindles, they will be forced to serialize access to the whole unit.
On the other hand, f you had three logical units in a RAID1 configuration, you will be able to have up to three concurrent accesses.
This does disregard the transfer and access rate differences between a single RAID1 and a RAID10, so it's possible it would come out a wash.
In a worst case, you could have SQL code selecting from tempdb, inserting into a data file, and logging for the inserts flowing into the transaction log. And this could all take place in a single SQL command.
So which would you rather have in this scenario? Concurrent access to 3 RAID1's or a single wide RAID10? I'm betting on the former.
BTW - Great blog Kendal!
-Graham Tapscott
Post a Comment