|
|||||||||
![]() |
|
Thread Tools | Search this Thread |
![]() |
#16 |
Major Player
Join Date: Jan 2006
Location: Birmingham, AL USA
Posts: 722
|
Windows can do drive spanning, and I think even RAID-5 (on the server platforms I know, but unsure of XP Pro). Hardware RAID is the best and fastest way to go.... here's a great web site to understand the different types of RAID
http://www.raid.com/04_00.html |
![]() |
![]() |
![]() |
#17 | |
Trustee
Join Date: Feb 2004
Location: Suwanee, GA
Posts: 1,241
|
Quote:
Folks, you always have a chance to lose data and the biggest threat usually sits in the chair. Have a plan and backup what you must have on alternate media. This is why you also keep your tapes. That is your backup on full disaster (machine related). Some NLEs will backup projects with media, but storage is almost impossible without big expense. RAID 0 is an ok way to get bigger volumes. RAID 5 is always better. And for the record, most RAID units are not faster, just more reliable. The wait though is measured in computer time and not yours and mine. You will not notice except on bad configurations. Warning - do not mix software RAID and hardware RAID. This will cause a massive slowdown. Software RAID is the last choice unless you want RAID 5, don't care about performance, and cannot get a hardware solution. |
|
![]() |
![]() |
![]() |
#18 |
Trustee
Join Date: May 2004
Location: Knoxville, Tennessee
Posts: 1,669
|
>The wait though is measured in computer time and not yours and mine. You will not notice except on bad configurations.
Respectfully, that has not been my experience. The improvement in my system's responsiveness was clearly noticeable (and measureable with benchmarkinng software too, although apparently benchmarks ain't 'real-world' enough for some folks) when I simply changed my drives to RAID 0 configuration. And as an aside, Cineform specifically recommends RAID 0 for use with their excellent Aspect HD product. |
![]() |
![]() |
![]() |
#19 |
Trustee
Join Date: Feb 2004
Location: Suwanee, GA
Posts: 1,241
|
Storage Reviews cover RAID performance.
|
![]() |
![]() |
![]() |
#20 |
Trustee
Join Date: Feb 2004
Location: Suwanee, GA
Posts: 1,241
|
http://faq.storagereview.com/tiki-in...leDriveVsRaid0
A link to the Storage Review FAQ. |
![]() |
![]() |
![]() |
#21 |
Inner Circle
Join Date: Sep 2002
Location: San Mateo, CA
Posts: 3,840
|
From George's reference link, the relevant statement;
"There are certain uncommon situations where RAID 0 can significantly improve system performance. For example, editing of large audio or video files is sometimes limited by the maximum sequential transfer rate of the hard drives, but it is far more common for the processor to be limiting factor. Generally, if you frequently make simple edits to large media files, RAID 0 can potentially improve your productivity. Examples of "simple" edits might include removing portions of an uncompressed audio file, or combining two video files. Most "filters" in media editing software are not bottlenecked by disk speed. Note that the audio and video editing that I am referring to does not include encoding digital music (such as Ogg/MP3 compression) or encoding/transcoding video (such as converting a DVD video file to DivX, etc.). These tasks are CPU limited, so RAID 0 would have little or no effect" So, RAID 0 does improve editing, but not compression tasks. |
![]() |
![]() |
![]() |
#22 |
Trustee
Join Date: Feb 2004
Location: Suwanee, GA
Posts: 1,241
|
Another thing, and this only really applies to big files, is the block size. When I created my RAID0 volumes, I used a 64k block size and then formatted those drives that way. I sacrifice 'slack' space if my files are small. But instead of reading a bunch of little pieces, I pull bigger chucks.
Slack space is the unused portion that is 'wasted' for each file I write. Say my block size is 1k. I write 1000 10-byte files. That is 990bytes wasted for each file (making this simple by rounding as a block really is 1024). For all 1000, that is 990k of 'wasted' space as each file must live in a single block. But, if I have a 1G file (we will use decimel and not binary again), that is 1,000,000k of data or 1 million 1k blocks. But if the blocks are 64k, there are only 15,625 of them. Of course, memory is 4k blocks and the system has to sort that out, but it can write to memory a LOT faster than read from disk. And the controller has to read the blocks, etc. Lots of variables, but R0 can be faster IF big blocks, formatted for it, and tuned. But for most mundane things, it is usually not significantly faster. |
![]() |
![]() |
![]() |
| ||||||
|
|