ph0rk Posted November 17, 2007 Report Posted November 17, 2007 It's pretty sad when the pictures from my digital camera and the porn I've downloaded off the 'net have more redundant backups than head-fi does. Telling, in a way. I am still amazed there weren't at the -least- on-site tape backups. How much does a tape robot cost these days? $3k? Don't back up images - done. C'mon! RAID is not a backup, even if you have a mirrored 5 array with a spare parity disk per array. Controllers fail or several well-placed drives go and bam. RAID arrays are for availability not real backups. I avoided slamming HF for the outage until now, but a week is just awful. But then, head-fi didn't have strikethrough, so it was clearly an inferior installation.
fordgtlover Posted November 18, 2007 Report Posted November 18, 2007 Well I think the problem your seeing in this case is that the live data and the backups were kept on the same NAS device this is a major no no. I'll assume in your case that you'd be a little smarter and not keep live data and backups on the same device . Hardware fails it's the nature of the beast which is why backups should always be on a different device and/or off-site depending on the importance of the data. 100% RAID is not a backup, even if you have a mirrored 5 array with a spare parity disk per array. Controllers fail or several well-placed drives go and bam. RAID arrays are for availability not real backups. It's quite simple - only keep offsite backups of the data you don't want to lose, or can't afford to lose.
elnero Posted November 18, 2007 Report Posted November 18, 2007 Jude's plans for world domination of the headphone world have been canceled. And yes payback is a bitch. The creator says not to gloat in the misery of others, but what jude has done to me and my friends, and the way he treats everyone is so condecending that he deserves this and much more. Jude needs to stop lying and look in the mirror for the source of all his trouble. He believed too much of his own bullshit. OK Kevin, enough beating around the bush, now tell us how you really feel.
Edwood Posted November 18, 2007 Report Posted November 18, 2007 Yeah, I'm kind of with Grawk - the best rebuild is a clean one from a virgin wipe. The problem is that Jude had so much crap built on top of things with all that advertising, there's probably no way to re-create things in a stable manner. The further away you get from the base forum software, the harder it will be to re-create it in a stable state. Did anyone notice that d*mn "Intellitext" that went on in the last few months? That stuff is a virus all on its own. Head-Fi Reloaded. Do not want.
yuujin Posted November 18, 2007 Report Posted November 18, 2007 Yeah I agree with ph0rk. RAID 5 is meant for data redundancy(if one or two hdd screws up) and never back up. @kevin EMC is still using consumer grade hdds for their lower end NAS.
kevin gilmore Posted November 19, 2007 Report Posted November 19, 2007 @kevin EMC is still using consumer grade hdds for their lower end NAS. Yeah, i know. Too much competition forces a crap low end product. Too many employees at EMC these days, going to have to downsize considerably. In a few years you are going to see lots of raid boxes built from solid state with ultra high speed side channel backup.
yuujin Posted November 19, 2007 Report Posted November 19, 2007 Are you serious? I have a friend working there. Thats bad news man!
Post Posted November 21, 2007 Report Posted November 21, 2007 Didn't Head-fi start because Headwize went down too? Don't tell me this is the passing of the torch-by-way-of-downs. Honestly I think Head-Case would rather stay lean and oh so meeeaaaannn and can't wait for Head-fi to get back up and running I love you both! Hugs and kisses!
itsborken Posted November 21, 2007 Report Posted November 21, 2007 Yeah I agree with ph0rk. RAID 5 is meant for data redundancy(if one or two hdd screws up) and never back up. @kevin EMC is still using consumer grade hdds for their lower end NAS. raid5 is cheap data redundancy. When one has an older array and the disks go marginal and fail, one can find all sorts of minor corruption after a parity rebuild. Enough minor corruptions can snowball into an unusable database table/application and complete rebuild. Has it been established that EMC was the guilty party?
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now