Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people, just like you, are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions

Hard Drive Speeds 7.2krpm VS 15krpm.

Posted on 2013-10-27
Last Modified: 2013-12-07
Would like EE opinions on the following. Looking to buy a NAS device to expand disk space drives available are: drives available Couple concerns, I've always purchased 15krpm drives in the past for performance however now the space I need doesn't make sense to purchase two or more servers, will using a 3T or 4T drive at 7.2rpms create a problem? I prefer to get the max amount of storage from one unit, the one I'm looking at holds 12 bays, obviously 1-4T drive is the better option concerned about the performance hit when it comes to cutting the rpm's in half.
Question by:WORKS2011
LVL 34

Expert Comment

by:Dan Craciun
ID: 39604806
If you're connecting to the NAS via gigabit ethernet, with a 12 bay NAS I think you'll saturate the link with both type of drives.

LVL 70

Accepted Solution

garycase earned 500 total points
ID: 39604988
There are two performance factors to consider ...

(1)   The rate at which data will transfer from the NAS.   As noted already, either drive will have a sustained data rate that can saturate a Gb Ethernet connection => so unless your NAS supports NIC teaming or 10Gb Ethernet, there's NO real difference between the drive for this.

(2)  How quickly a transfer will start after requested.   15k drives have notably better access times than 7200rpm drives do.   In addition to better seek times (typically 2/3rds that of a 7200 rpm unit), they have 1/2 the latency times ... the net result is a much higher transaction rate.  

Bottom line:   If you do a LOT of small transfers, there will be a significant advantage with the 15k drives.     But for most purposes, you simply won't notice any difference -- and clearly you'll have FAR more storage with the slower, much higher capacity drives.
LVL 42

Expert Comment

ID: 39604989
More important than the drive RPM speed is the RAID topology. At least disk write operations are much slower on RAID 5 than RAID 10. RAID topology, of course, has a big impact on the total available storage capacity...

The last but not least option to think about is the server RAM memory size. We have 256 GB and almost all frequently used data are buffered in RAM so the RAID speed is not the main concern.
U.S. Department of Agriculture and Acronis Access

With the new era of mobile computing, smartphones and tablets, wireless communications and cloud services, the USDA sought to take advantage of a mobilized workforce and the blurring lines between personal and corporate computing resources.

LVL 47

Expert Comment

ID: 39610829
In the grand scheme of things in a server, RPMs are about as relevant as the brand of mouse you use.

SAS vs SATA; RAID level; chunk and raid stripe I/O sizes;  I/O profle (reads vs writes vs seqential vs random); host buffers & cache; HDD cache/buffers;  power saving; redundancy level;  bus controller saturation; peripheral saturation;  load balancing & multipathing …

All of those can have a much higher impact than RPMs, and thy all are affected by a dozen more parameters .
LVL 70

Assisted Solution

garycase earned 500 total points
ID: 39611160
"... RPMs are about as relevant as the brand of mouse you use. "  ==>  Definitely do NOT agree.     One of the key factors in a storage array is IOPS, and the 3 main factors in how many IOPS a drive can are are  seek time, latency, and transfer rate.

The first two are directly related to rotational speed of the drive => faster rpm = quicker seeks and lower latency.    The 3rd is also related to the rotational speed, but depends also on the areal density of the platter = i.e. given the same areal density a higher rpm drive will be faster;  but many 15k drives use lower areal density platters, so this may be a wash.    Also, as I noted earlier, the smaller the average transaction, the bigger the benefit is with the higher rpm drives, since if you're doing a lot of small transfers, the impact of the seek time and latency is much more important than the transfer rate.

In addition, the "RAID penalty" that slows down writes (due to multiple disk operations), while numerically the same factor [e.g. 2 operations for RAID-1 or 10, 4 for RAID-5, 6 for RAID-6, etc.],  clearly the more IOPS the system can sustain, the more "net write IOPS" it
can achieve in environments with a high % of writes.

If the number of IOPS needed is small, then I certainly agree that slower drives are fine ... and for many applications that's true.    But it's certainly a factor that should be considered in choosing the characteristics of the drives.
LVL 47

Expert Comment

ID: 39611578
I was trying to reveal a counter-intuitive truth when it comes to RPMS.  They really aren't all that important in a RAID system, and people who are trying to build a storage array put way too much emphasis on RPMs when they should be looking at so many other things that will have a much larger impact on overall storage performance in the real world.

So here is a thought experiment to prove why RPMs are incorrectly viewed as being more important than they really are.
*  Why will a pair of 7200 RPM disks in RAID1 significantly outperform a single 15K RPM drive in just about any benchmark on the planet?  
* Why will the SAS version of a seagate constellation drive outperform the SATA version of that same drive (which has same RPMs & cache)?
*  Since one can saturate an expander with 2 or 3 disk drives anyway in high throughput situations, then how does the RPMs make a difference once the bus is saturated anyway?

Certainly if you are looking at a single HDD in a contrived benchmarking situation, RPMs are going to be highly relevant …. until the data is loaded in the cache buffer of the drive.   Then the performance is gated by the bandwidth available on the bus and the physical interface.

As more disks are added and you have load balancing and controllers cache I/Os and you have to start waiting for a window of opportunity to transfer data across the physical interface due to bus limitations, then the impact of RPMs become less and less interesting when viewing performance of the storage subsystem as a whole.

RAID levels,controller choices, SAS vs SATA, vibration, caching, filesystem settings all contribute much more to performance as a whole than RPMs ever will.

Featured Post

Portable, direct connect server access

The ATEN CV211 connects a laptop directly to any server allowing you instant access to perform data maintenance and local operations, for quick troubleshooting, updating, service and repair.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

VM backups can be lost due to a number of reasons: accidental backup deletion, backup file corruption, disk failure, lost or stolen hardware, malicious attack, or due to some other undesired and unpredicted event. Thus, having more than one copy of …
Learn about cloud computing and its benefits for small business owners.
This video Micro Tutorial explains how to clone a hard drive using a commercial software product for Windows systems called Casper from Future Systems Solutions (FSS). Cloning makes an exact, complete copy of one hard disk drive (HDD) onto another d…
This tutorial will walk an individual through setting the global and backup job media overwrite and protection periods in Backup Exec 2012. Log onto the Backup Exec Central Administration Server. Examine the services. If all or most of them are stop…

791 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question