If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#121
|
|||
|
|||
flux writes:
Hopefully, mostly in RAID boxes attached to those servers. Certainly drives are the most common component to fail in systems. That's true Actually, it's the LEAST likely. No, wait, I take it back. Power supplies and fans are the LEAST likely to fail. See http://www.bluemax.net/docs/Motherboards/General_Info/mtbf.html -- The components that experience the highest failure rates are [...] keyboards, mice, disk and tape drives, monitors and power supplies. But I suppose a PC manufacturer wouldn't know. ;-) SCSI product life is longer, allowing longterm field reliability I think the point is that they aren't. Technically, this is "product life", e.g. how long a particular model is manufactured. I think you haven't said anything about that. So maybe SATA drives can take heat better than SCSI drives? *laugh* It seems make sense that more drives you have, the more chance for failure. With each extra drive, that's a whole other set of electronics, actuators and spindle motor that can fail. Half right. There's more chance of an individual drive failing, but less chance of data loss, in a RAID constructed of smaller disks (given a fixed size of the RAID and fixed parity grouping). Your math is backward. Which part? ;-) For RAID 5, the mean time to data loss is MTTF(disk)^2 ---------------------- N * (G-1) * MTTR(disk) Spare us. :-) -- Anton |
#122
|
|||
|
|||
In article , Anton Rang
wrote: Actually, it's the LEAST likely. No, wait, I take it back. Power supplies and fans are the LEAST likely to fail. See http://www.bluemax.net/docs/Motherboards/General_Info/mtbf.html -- The components that experience the highest failure rates are [...] keyboards, mice, disk and tape drives, monitors and power supplies. But I suppose a PC manufacturer wouldn't know. ;-) At least not one apparently makes junk. SCSI product life is longer, allowing longterm field reliability I think the point is that they aren't. Technically, this is "product life", e.g. how long a particular model is manufactured. I think you haven't said anything about that. I don't get it. longterm field reliability = product life? So maybe SATA drives can take heat better than SCSI drives? *laugh* You can't have it both ways. They either stand the heat better or are not as hot. Your math is backward. Which part? ;-) That was it. |
#123
|
|||
|
|||
In article , Anton Rang
wrote: Mark Hahn writes: They tested 300 drives of two types, SATA and FibreChannel, under 1000 continuous hours of three workloads. (The workloads were derived from hmm. MTBF of current drives is around 1M hours, and they tested .3M hours of operation. I don't really see why this is a meaningful sample, especially considering the standard "bathtub" curve for expected failures. I think the interesting bit is the comparison between failure rates under different load, more so than the absolute values. That's if you believe their results. And at best it just tells us that Seagate has problems with *their* SATA drives. In any case, we don't know what criteria are for the different loads, at least according to these slides. For example, does "High Duty Cycle" imply "High End Server"? The other two refer specifically to a type of computer usage, but "High Duty Cycle" doesn't say where that usage occurs. For that matter, it's not really clear why random access is considered to a be "Low End Server". Seems they both could either low or high end depending on the actual usage. Seems to me that "High Duty Cycle" would be something a Tivo. But even ATA drives in Tivos seem to work great. I guess they must not be using Seagate drives. |
#124
|
|||
|
|||
In article ,
flux wrote: usage. Seems to me that "High Duty Cycle" would be something a Tivo. But even ATA drives in Tivos seem to work great. I guess they must not be using Seagate drives. I suspect that a drive in a Tivo would have a pretty atypical access pattern, assuming they take advantage of the fact that they're almost always streaming data into or out of the disk. Which I would bloody well hope they would do. This should mean massively less seeking than just about any server workload. -- I've seen things you people can't imagine. Chimneysweeps on fire over the roofs of London. I've watched kite-strings glitter in the sun at Hyde Park Gate. All these things will be lost in time, like chalk-paintings in the rain. `-_-' Time for your nap. | Peter da Silva | Har du kramat din varg, idag? 'U` |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Seagate Warranties Jump to Five Years | Ablang | General | 0 | August 1st 04 02:43 AM |
Seagate Redesigns Drives (with 73GB to 300GB capacities) | Ablang | General | 0 | May 23rd 04 04:01 AM |
Western Digital, Maxtor or Seagate | @drian | Homebuilt PC's | 26 | October 20th 03 06:24 PM |
Western Digital, Maxtor, Seagate - guess who spams? | tiburón | Homebuilt PC's | 7 | September 29th 03 11:19 PM |
My Maxtor saga | Steve Daly | Storage (alternative) | 25 | August 4th 03 04:12 AM |