wparsonsgisnet Posted April 27, 2007 Share #41  Posted April 27, 2007 Advertisement (gone after registration) Bill when you say UPS - is that surge protector? So it seems that it's safer to keep them powered up all the time? The LaCie I bought seems to work fine so far - I'm planning to buy a second 600gb from them refurbed. Jonathan Elderfield Photography  Jonathan, UPS is uninterruptible power supply, that is, battery backup. I have some with surge protection built in, but I'm not sure about the one that I have idle at the moment.  In any case, I use multiple surge protection in a series all the time. These things are cheap and at least one in the chain is going to work.  a friend of mine, whom I respect greatly, says that UPS is the opposite of DOWNS. Link to post Share on other sites More sharing options...
Advertisement Posted April 27, 2007 Posted April 27, 2007 Hi wparsonsgisnet, Take a look here External Hard Drive Recommendations. I'm sure you'll find what you were looking for!
foolstop Posted April 27, 2007 Share #42  Posted April 27, 2007 Bill, thank you for your compliment  Google has their own Google (global) file system (GFS). I doubt whether they need to run RAID5 in any of their computing nodes, RAID 0 maybe, for performance sake. Instead, they make the entire computer (node) redundant.  This is a very interesting article concerning the data management of Google. How Google Works - How Google Works: The Start of the Story  "The computer components going into those racks were also purchased for rock-bottom cost, rather than reliability. Hard drives were of a "poorer quality than you would put into your kid's computer at home," Merrill says, along with possibly defective memory boards sold at a fire-sale discount. But that was just part of the strategy of getting a lot of computing power without spending a lot of money. Page, Brin and the other Google developers started with the assumption that components would fail regularly and designed their search engine software to work around that."  The idea is to "store data reliably even in the presence of unreliable machines,"  ===================== In one your Lacie failure case, only the enclosure electronics has failed. Not the drive itself, that would be no problem at all, you can bought a new enclosure with Firewire with USD around 30 nowadays. In case of drive electrical or mechanical failure, I am quite sure the data stored in the magnetic platter can be retrieve (with a high cost), in that sense, harddisk is a good backup device, even compared with DVD+R archive.  In general, I only use Seagate drive, the ES series of drives (only available in SATAII interface), are supposed to run in 7x24  The google disk failure analysis article also mentioned the "infant mortality" problem, so be sure to test drive the disk for a period before you start to rely on this particular disk. =====================  I am connecting an UPS to my new RAID, in general, they don't draw a lot of power, the UPS should be able to give enough time to signal the RAID and power it down. BTW, quite a no. of these RAID solutions, Infrant or Buffalo, use ext3 as the file system, it is a journaled file system, should be able to add further protection to system power failure.  Concerning surge protection, last summer, I added an UPS to my TV when I was going to watch the World Cup Football in a evening with heavy thunderstorm  Nevertheless, some surge damages are due to huge surge going thro' the data/phone lines or the "ground loop", that could not be protected by the UPS, so would be nice if one able to unplug all the lines from the wall during heavy thunderstorm hours. Link to post Share on other sites More sharing options...
wlaidlaw Posted April 28, 2007 Share #43 Â Posted April 28, 2007 Having read even the limited sample in this thread, I am not convinced RAID is the answer due to the high failure rate of the drives. Before I retired, for business I used an IBM system similar to RAID on our IBM mini servers. The hard drives failed regularly and due to the age of the equipment, it became very difficult to find new ones. I resorted to buying spare hard drives on eBay and getting them refurbished. If you looked at the drive statistics, you found that even during the night, they were reading and writing to each other. I came to the conclusion that the RAID type of set up means that the drives work much harder than non-RAID. I found some different software for our minis that changed from RAID to a mirrored set up and our hard drive failures dropped dramatically. Â Whatever you buy, make sure it is properly cooled. In France I have a solid alloy cased 250GB Maxtor made by Connectland. It gets too hot to leave on continuously, which means I have to use the unsatisfactory alternative of switching it off and on. This is obviously not going to work with Time Machine once Leopard comes out and I will have to replace it. I can then just use it as a secondary back up unit. On my Powermac in the UK I have 2 x 250GB internal SATA drives, which I use mirrored with Chronosync. Â While I still had a windows server at home in the UK, I had a Netgear SC101 network drive with 2 x 160 GB EIDE drives in it, plugged into the router. The hardware is fine on this machine but both the firmware and software stink. If you mess up the drive partitioning allocations for various family members' back-ups, it is virtually impossible to undo. I had to resort to wiping the drives in DOS, plugging them into the main drive bay of my PC and reformatting them. It is now sitting in a cupboard, as apart from anything else, Netgear never brought out the Mac driver for it that they said they were going to, when I bought it (Mac driver due soon it said on their website in 2004). Â Wilson Link to post Share on other sites More sharing options...
stunsworth Posted April 28, 2007 Share #44 Â Posted April 28, 2007 Wilson, Raid and mirroring achieve different things. With a most Raid arrays if a drive fails you can simply replace the failed drive with a new one and continue as if nothing had happened. That isn't the case with mirroring where at the very least you to switch over to the mirrored drive and start using that. Not too difficult if it only contains data, but a bit more complicated if you are mirroring applications and/or the OS. Â I mirror the data on my 1 TB external drive, but that's two thirds full and I'm considering moving over to a Raid box with the photographic folders mirrored to another box. Link to post Share on other sites More sharing options...
wlaidlaw Posted April 28, 2007 Share #45  Posted April 28, 2007 Wilson, Raid and mirroring achieve different things. With a most Raid arrays if a drive fails you can simply replace the failed drive with a new one and continue as if nothing had happened. That isn't the case with mirroring where at the very least you to switch over to the mirrored drive and start using that. Not too difficult if it only contains data, but a bit more complicated if you are mirroring applications and/or the OS. I mirror the data on my 1 TB external drive, but that's two thirds full and I'm considering moving over to a Raid box with the photographic folders mirrored to another box.  Steve,  I agree that RAID is in theory, much more convenient than mirroring, if it were not for the apparent hard life the hard disks have. It is like in theory, Firewire (especially 800) is much better for the connection of an external hard drive than USB2 but if you look at the problems people have with it (hard discs refusing to mount seems to be the main beef), I will stick with USB2 or peer to peer Ethernet.  Wilson Link to post Share on other sites More sharing options...
foolstop Posted April 28, 2007 Share #46 Â Posted April 28, 2007 Wilson, I suppose RAID 1 or RAID 5 is actually the strategy to address the drive failure problem. Anyway, some form of disk redundancy is the way to go. Â Steve, have you ever encounter the infamous "XP USB delayed write error"? Chaos Manor Reviews - Mailbag for January 2, 2007 Â It does occur rarely, but the consequence however, is serious. One of my friend ran into the delayed write problem during one of the backup, and he did not notice it sadly, until weeks later... Â So one must compare the backup and the original after each copy. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.