I've now successfully set up two RAID 1 arrays on an ASUS P4C800 Deluxe motherboard...one pair is SATA and the other is ATA133. I'm hoping that this will provide some security for a Gigastudio rig that's meant to be a performance instrument rather than a studio instrument. I'm a bit confused about some details.
The common advice is that the 2nd drive needs to be bigger than the first (or better yet, that both drives be identical). But if I do the following:
- Use a 60 Gb and an 80Gb drive to create a RAID 1 array.
- Use a 60 Gb drive to create a replacement drive in case one of the RAID drives fails.
Now, if the 60Gb drive fails. I'm left with an 80 Gb drive still configured in the Array. But if the rule is that you need to use a drive that's the same size or larger to create an array, can I now put in the 60 Gb drive spare to replace the failed 80 Gb drive.
Maybe the limitation only applies when building or rebuilding an array? Maybe one could actually use a smaller drive as the second drive in an array, but it's discouraged because then you could write stuff to the first drive but it wouldn't get copied to the second drive if you get to the point that you've filled the smaller drive?
And another question. Why do several people in this forum think that it's more important to use raid for the sample drive than for the program file in the case of Gigastudio. I would think that, on stage, it's equally important to use RAID for the program file.
And finally, since I have several laptop drives available, I'm using 5400RPM drives for the system drive RAID with Gigastudio on it. It seems to run nicely. Am I asking for trouble using such a slow disk for the Giga program? I think the laptop drives may be able to handle more punishment than 3.5 in drives.
Drive size is inconsequential, but you obviously cannot hope to get 80GB's of content from a 60GB HDD.
I prefer a hot swappable copy personally. The 2 seconds it takes to swap them is fine with me, and I have never needed to do that anyway. But chance favors the prepared mind.
Onboard RAID will unnecessarily eat CPU cycles, and a RAID card is the best solution, but it also clogs up the bandwidth.
Unless it's a Brittany Spears production using pre recorded tracks, it's uneeded IMHO.
Did you happen to see Shania Twain and her fake, weak, band fall on their faces at the Superbowl 2 years ago?
The laziness that occurs when you repeatedly fake your gig is detremental IMHO, but it seems like that's where many " performers " are at these days. Sure, Cubase kept playing as told, but the human factor always gets you. Sting and Gwen Steffani followed her pitiful performance, and showed her how it's done on the big jobs. Live is where it's at for me. But I do use hardware sequencers in a production that I do weekly, while I play live down the street. So it does have it's uses. But that gig has an A/B switcher where 2 DAW's are run side by side. That's the kind of RAID I like.
I am sure you will use this live, and you will see that these apps, and the hardware will do just fine. But the less pipes you have in the plumbing the better IMHO. Try a simulated crash and see what the results are. RAID 5 will kill you while it rebuilds, and RAID 1 might let you finish as the spare takes over, but there is still an audio dropout when the HDD's switch over, for an extra second, you could run hot swappable spares, and only you and I might notice the dropout.
My partner former owner of Ultimate Ears, is Eddie's FOH man. He runs a digitally controlled analog console, and doesn't need automation except for group bussing, etc. But their Key Tech decided to try and play the fake keyboard tracks @ a different sample rate, and caused a major trainwreck in front of 1000's of fans. Another example of hardware doing what it's told, and humans failing. This turn the switch off and on jive causes mental laziness, and a false sense of security, which results in catastrophy at a live gig.
I must say, I often loathe the results.
Listen to Eddie and his son searching for the " new sample rate " that the keytech chose. A quick tuning of 3 quarter tones would have helped, on all 6 strings. 48Khz vrs. 44.1KHz, that's a freakin' nightmare.