-
Xserve Raid & FCP
We just switched from Avid to FCP (4.5) last January. However, we went with Huge Media System’s Media Vault, and it has consistantly been crashing, resulting in loss of media and uncounted days of downtime and increasing levels of frustration. It has been, in short, a nightmare.
We are seriously considering investing in Apple’s Xserve raid. But before we plunk down $6000 (for their cheapest version at just 1TB) We would like to hear from others, if any, using this configuration and research any issues they are having.
We are using a dual 2.5GHz G5, with NVIDIA GeForce 6800 GT DDL (takes up adjoing PCI slot, besides the AGP slot), and a SCSI card, which we will need to keep to write to our DLT tape drive when mastering DVDs, so we have just one more slot open for the necessary fibre channel card.
Any advice, commiseration, encouragement and comments appreciated!
-
8 Replies
-
Ours has been running for 1 year now with no failures and no glitches what so ever. Rock solid performance. We are running a G5 dual 2.0 w/ 4 gig ram and 3.5 tb XServe raid along with an AJA I/O to input uncompressed Digi Beta and DVCPro via SDI.
I like it so much we are upgrading our 2nd suite from Medea drives to a new G5 and an XServe. I could save some money, but reliability and performance are hard to put a price on. I want to know my equipment is not impeding the creative process, which is what good editing comes down to.
-
I can help with the encouragement … we’ve been using Apple’s Xserve RAID systems and G-TECH’s 500GB G-RAID’s since January with zero issues and no complaints so far.
wishing you luck with your purcahses
Andy -
On the flip side, have a contacted Huge? They have awesome support and I’m surprised to hear they didn’t work it out with your Media Vault. I understand your pain as I went through the same frustrations with a SATA RAID, but I only had myself to support it. Ever since I switched to Huge I have had no major problems. I had a little glitch and they promptly told me how to fix it. What format video do you mostly work in? What RAID level was your HUGE set to?
Sorry to bug you about your frustrating circumstances, but I’m just curious.
Jeremy
———–
G5 Dual 2Ghz <> 4GB RAM <> FCP 4.5 <> Kona 2ATTO 42XS <> Huge Systems 1.25 TB 4105 Fibre
-
We just switched over to an Xsan from an Xserve RAID. With the RAID I had some problems, mostly with dropped frames during digitization (SDI, Dual 2G G5, etc.) and inconsistent insert edits with ETT (which may have been FCP 4.5 related). Initially, we had severe sync issues that traced back to the RAID, and toward the end I had to digitize all my 10-bit stuff to a local drive and then transfer it. The RAID wouldn’t allow any 10-bit material to import without dropping frames or stalling the batch capture. It may have been solvable, but once you’ve got 600GB of stuff already on it it’s a little diffiicult to troubleshoot without losing it all. So, since we we’re expanding anyway, we just set up an Xsan network and sold the RAID.
Judging from what everyone else has said about the Xserve, I think mine might have been cursed.
-
Jeremy,
It is true that I cannot fault Huge’s customer support, which has been exlempleary. However, since last Easter, the Huge Media vault has failed completely at least half a dozen times. Disk Warrior has been ineffective in saving any useful data, finding and fixing files on the first pass, but consistantly failing to write back the fixed files to the raid, resulting in total loss of the data, and requiring a complete reformat of the drive. (If anyone wants the steps, I can, at this point, recite them from memory)
In fairness, we discovered that at least some of the problems are a result of an “undocumented application behavior, ” which we used to call a “bug” in FCP 4.5 (since fixed, I’m told, in 5.0)in which FCP allocates the entire empty drive space during capture now. After capture, FCP 4.5 would then have to de-allocate the rest of the drive, and if you have several hundred gigs open, this takes several hours. Force quitting during de-allocation corrupts the raid, resulting in catastrophic data loss.
We are working with 10 bit uncompressed Beta and DVcam footage for the most part, injested via AJA IO.
We are working with the Media Vault in redundant mode, which gives us approx. 500-600 GB of space, that is, when it is working.
-
I’ve been running a 2.5Tbyte xserve raid for 1 yr. It’s stripped as one big volume so i can do uncompressed HD. I bought the apple care protection plan and i’m glad i did. I had a drive go bad and they replaced it overnight. The cool part was that even though it’s one big drive I didn’t lose any info. I have had occasional drop frame error messages on playback, but I’ve never been able to isolate it to the xserve and they are so infrequent, they’re not a serious issue. It was a spendy decision, but I’m glad I’ve got it.
-
We’ve been running 45 terrabytes of X-Raid storage using X-SAN for the past year and a half. So far, only one drive failure during that time. Because we were using Raid 50, we simply popped out the bad drive and popped in a new one. No loss of data. No downtime. Not the cheapest solution, but in my opinion, the best.
mark
-
I am having the same problems regarding capturing 10 bit and dropping frames going to the RAID. I am using a blackmagic decklink pro HD with dual link. I also have the dual 2.5 G5 and am running FCP 4.5 on Tiger OS. Have you found a solution yet?
Thanks,
Bear
Reply to this Discussion! Login or Sign Up