Question about External HD
in [Mac Systems]
|
Prev: X11 startup configurations
Next: Secure Empty Trash
From: David Empson on 25 Jan 2010 19:44 nospam <nospam(a)nospam.invalid> wrote: > In article <4b5e0d79$0$8568$ba624c82(a)nntp06.dk.telia.net>, Erik Richard > S�rensen <NOSPAM(a)NOSPAM.dk> wrote: > > > It is you who began talking about harddisks and SCSI - not me. I just > > pointed it out to you that you can't compare anything like that with USB. > > i cited the 8mhz mac because back then, the cpu *did* matter. a 2.5 > gigahertz core 2 duo is not going to be burdened by a couple of hard > drives, especially when they are limited to usb speeds. > > > The 'problem' is that the USB still is a sort of serial construction > > where each port is depending on the other ports. > > nope. > > > There are simply no Mac computer with built-in independant USB > > circuitry. > > every mac since the dual usb ibook has independent usb ports. that's > why it was called a dual usb. That iBook was called "Dual USB" because it was the first iBook with two USB ports. It also had a dual root hub, but that arrangement existed earlier in other models. From Apple's developer note for the Dual USB iBook: [begin quote] USB Interface The Pangea IC implements two independent USB controllers (root hubs), each of which is connected to one of the ports on the back panel of the computer. The use of two independent controllers allows both USB ports to support high data rate devices at the same time with no degradation of their performance. If a user connects a high-speed (12 Mbps) device to one port and another high-speed device to the other, both devices can operate at their full data rates. [end quote] Almost all Macs with at least two USB ports have at least two USB root hubs. I can confirm this from developer notes for the following models: PowerMac G4 (AGP Graphics) - mid 1999 iMac (Slot-Loading) - late 1999 PowerBook G3 (Firewire) - early 2000 iBook G3 (Dual USB) - mid 2001 Earlier tray-loading iMacs and the PowerMac G3 (Blue & White) had a single USB controller and single root hub with two ports. I expect the PowerMac G4 (PCI Graphics) is the same. For later models, it doesn't quite extend as far as every built-in USB port having its own root hub: there are 17" MacBook Pros with three USB ports but only two root hubs. I observed this with at least one model which has two USB ports on one side and one on the other. I think I noticed a similar arrangement with some iMac models that had three or more ports. There may be others but I haven't done an exhaustive search (and Apple have stopped publishing developer notes so I have no idea for models in the last couple of years). Some internal USB peripherals are attached to the same USB hubs as the external ports, but some have an independent root hub. For example, my MacBook Pro has the built-in iSight on the same high-speed hub as the right USB port, and built-in Bluetooth on the same low-speed hub as the left USB port, but the keyboard/trackpad and IR receiver are on a separate low-speed hub. The arrangment can be observed in System Profiler for a particular Mac (you need to plug a device into the ports to work out which one is which). > > > they're both serial as far as the cabling goes, but you could say that > > > firewire is more 'serial' since it is daisy chained, whereas usb is > > > more 'parallel' since there are multiple devices off one hub. > > > > For that manner you can also 'daisy-chain' USB devices, > > you absolutely cannot daisy chain a usb device to another device. > > only if there's a built-in hub can you attach another usb device, such > as with a keyboard, but that's not daisy chaining, that's going through > a hub, one that happens to be included with the device. also, they're > usually bus-powered so the attached devices need to be low power. This arrangement will also be lower peformance, since a USB hub has to buffer and then retransmit entire packets, doubling the time required to transfer each packet. > > but that doesn't change anything since the transferred signal still is > > routed through the cabling, which the FW signal isn't. > > firewire signals are not routed through cables?? so the cables are just > for decoration?? Firewire devices are all on a common bus (shared between all built-in ports on the computer, including both Firewire 400 and 800 ports), with available bandwidth shared between them. If you have two 3.5" hard drives actively transferring data via a computer on Firewire 400, they will get roughly half the throughput (each) that either one would get if it was the only active device on the bus. (On Firewire 800 the same applies if the drives are fast enough to saturate the bus.) If you are copying data between two external 3.5" hard drives, and can't use Firewire 800, you will probably get best throughput with one being on Firewire 400 (connected anywhere in a Firewire chain) and the other on USB (plugged directly into the computer). Furthermore, each Firewire device in the chain will introduce a very slight delay (one bus cycle, I think), due to buffering and retransmitting signals received on one port and sent out the other port. The delay due to Firewire daisy chaining is nowhere near as bad as USB hubs. (Firewire hubs may cause similar "entire packet" delays, but I've never even seen one so I haven't tested it.) > > FW here is more like a cabled 'radio transmitter', where you send > > multiple signals with slightly shifted carrying signal frequencies > > to/from each FW device. > > this just gets even more bizarre. where did you read *that* ? Erik seems to have completely wrong information about both USB and Firewire. The key reasons that Firewire 400 gets better performance than USB 2.0 (despite USB 2.0 having a higher peak data rate of 480 Mbps) are: 1. Protocol overhead. USB requires more mucking around than Firewire. 2. CPU involvement. The USB protocol requires a lot more CPU involvement than does Firewire. USB transfers require heavy use of the CPU, which limits available CPU time for other tasks on the computer and may reduce USB throughput if it the CPU needs to be busy for other reasons. Firewire requires very little CPU involvement - a lot more work is handled by the controller (and it makes better use of direct memory access). 3. Bus bandwidth wasted by polling idle devices. Since USB is an entirely polled protocol, the CPU has to regularly communicate with every connected device. This limits peak bandwidth for any device when there are other devices on the same USB bus. Firewire is a multi-master protocol and idle devices use significantly less bandwidth (probably none except while the bus is being reconfigured, and occasional integrity polls to make sure everyone is still there). 4. The requirement of using USB hubs to connect additional devices, and delays introduced by them. -- David Empson dempson(a)actrix.gen.nz
From: Jolly Roger on 25 Jan 2010 20:36 In article <C7836101.512C4%nicknaym@_remove_this_gmail.com.invalid>, Nick Naym <nicknaym@_remove_this_gmail.com.invalid> wrote: > In article jollyroger-D98938.19310324012010(a)news.individual.net, Jolly Roger > at jollyroger(a)pobox.com wrote on 1/24/10 8:31 PM: > > > In article > > <nospam.m-m-2929A4.11364324012010(a)cpe-76-190-186-198.neo.res.rr.com>, > > M-M <nospam.m-m(a)ny.more> wrote: > > > >> In article <jollyroger-65A1E0.10053024012010(a)news.individual.net>, > >> Jolly Roger <jollyroger(a)pobox.com> wrote: > >> > >>> In my book it's Seagate > all others. Of course that's just my opinion, > >>> and you know what they say about opinions. : ) > >> > >> I have been reading that the Seagate 7200 2.5" drives have some serious > >> reliability issues that may or not have been fixed with a firmware > >> update: > >> > >> http://www.amazon.com/Seagate-Barracuda-ST3320613AS-Bare-Frustration-Free > >> -Packaging/product-reviews/B001VKYA5Y/ref=cm_cr_pr_hist_1?ie=UTF8&showVie > >> wpoints=0&filterBy=addOneStar > > > > You can find similar reports about just about every hard drive > > manufacturer in existence. All manufacturers make mistakes, and all of > > them have bad batches or drives from time to time. It's the nature of > > the business. Seagate drives are certainly among the most reliable in my > > experience. > > Seagate has had a slew of problems these past few years, as evidenced by > posts in its forums as well as user reports on retailer sites like Newegg. All drive manufacturers have had similar problems. There is nothing outstanding about Seagate in particular that I can see. > Though not alone in this regard -- overall, drive quality has, IMO, > plummeted because of Industry cost-cutting measures that have resulted in > increasing reliance on economically distressed offshore suppliers -- I have > the sense that Seagate has suffered the most. I don't have that sense at all. > (As you know, JR, the company > recently slashed its 5-year product warranty to 3 years, using the > transparently lame excuse that it was doing so simply to be "more in line > with the rest of the industry.") They reduced warranty coverage only on particular models: <http://achsyd1.achieva.com.au/cgi-bin/estore/news_20081215.html> It wasn't an across-the-board thing. > The other thing that might be worth mentioning is that Seagate's support for > Mac is (IMO) not as good as it is for PCs, as evidenced by its lack of a Mac > version of its HD diagnostic/utility software (I forgot the name). Show me the Mac utility for Western Digital drives. How about the Mac utility for Hitachi drives - where can I find it? Seagate is not outstanding in this regard. ; ) > Additionally, when I got my Barracuda (September, 2008), I was told by a > Seagate tech that they won't deal with any problems that occur if the drive > is used in an external enclosure (which, of course, is how I'm using it) -- > they don't want to get caught up in a finger-pointing contest between > themselves and the enclosure vendor should a drive problem arise. I would think that would be the position of most hard drive manufacturers. I don't see Seagate as outstanding in this regard either. My life experience has shown that Western Digital 3.5-inch drives are more apt to fail within warranty than Seagate 3.5-inch drives. And my life experience has shown that Seagate 3.5-inch and 2.5-inch hard drives are fairly reliable, and mostly last longer than the warranty period. In many years purchasing Seagate almost exclusively, I have only had two drives fail - one within warranty, and one a couple years out of warranty. Customer service for the drive that failed within warranty was exceptional. I had a replacement drive the same week, and had the option of getting the replacement drive before I sent the bad drive to them. I'll continue buying Seagate until my life experience changes. -- Send responses to the relevant news group rather than email to me. E-mail sent to this address may be devoured by my very hungry SPAM filter. Due to Google's refusal to prevent spammers from posting messages through their servers, I often ignore posts from Google Groups. Use a real news client if you want me to see your posts. JR
From: Erik Richard Sørensen on 25 Jan 2010 21:05 nospam wrote: > In article <4b5e3331$0$4810$ba624c82(a)nntp02.dk.telia.net>, Erik Richard > Sørensen <NOSPAM(a)NOSPAM.dk> wrote: > >> Sure it is. And sure it matters how many UsB devices you connect. It is >> obviously that you've never tried running with more than 1-2 maybe three >> USB external devices at a time. > > i have 7 usb devices on my desktop mac. Connected how? I guess not individually - rather by hubs... And you donø't have any slow down in speed when transferring to/from multi-disks? - If you claim that it is a veeeery special Mac you're using. I can guarantee you one thing and that /is/ that the speed is slowed down. I have only 3 USB devices connected, keyboard, trackball and scanner. And it's for sure that the speed is slowed down when the scanner is working at the same time as typing in a document! > >>>> There are simply no Mac >>>> computer with built-in independant USB circuitry. >>> every mac since the dual usb ibook has independent usb ports. that's >>> why it was called a dual usb. >> I have both worked with and seen rather many 'splashed' iBooks and >> PowerBooks for that matter - and none of these have had anything but 2 >> USB ports and that's why it is called the 'dual USB' model. > > from the developer note for the ibook dual usb: > > <http://developer.apple.com/legacy/mac/library/documentation/Hardware/De > veloper_Notes/Macintosh_CPUs-G3/ibook08May01/ibook-17.html#25515> > > The Pangea IC implements two independent USB controllers (root hubs), > each of which is connected to one of the ports on the back panel of > the computer. The use of two independent controllers allows both USB > ports to support high data rate devices at the same time with no > degradation of their performance. If a user connects a high-speed > (12 Mbps) device to one port and another high-speed device to the > other, both devices can operate at their full data rates. > > read that last sentence a couple of times. Yeah and? - Here's explecitely mentioned _two_USB_controllers_ not _one_ controller with independant circuits or one controller with dual or triple circuitry which is the normal on most common PCs. But even when there are those two controllers, you will experience some slow down in speed, since both controllers are relying on the same CPU. But it's also true that the slow down will be smaller in this construction. >>>> For that manner you can also 'daisy-chain' USB devices, >>> you absolutely cannot daisy chain a usb device to another device. >>> >>> only if there's a built-in hub can you attach another usb device, such >>> as with a keyboard, but that's not daisy chaining, that's going through >>> a hub, one that happens to be included with the device. also, they're >>> usually bus-powered so the attached devices need to be low power. >> Aha! Why can I then just connect one USB harddisk to the other, when >> there are both an USB A and an USB B port? - Nothing mentioned anywhere >> about built-in 'hubs' or things like that. > > which usb hard drive has both types of ports? The French brand, Archos, some of the Mercury enclosures from OWC and some of the Weiland professional series enclosures. >> - I also /have/ two harddisks >> with built-in hubs, but here it's explecitely mentioned that this a hub. > > exactly. if there's a usb a port, then there's a hub inside. > >> These UsB ports can even be used, if the drive is connected via FW400, >> FW800 or eSATA. > > i've never seen a drive that can use both usb *and* firewire at the > same time. some even have physical switches, although most auto-detect. > which usb drive is this? Enclosures from NewerTech. >> It is also clearly written in the guide that the USB >> ports are _balanced_ with _individual_ powering, but that doesn't mean >> that they are independant circuits. > > so what does it mean? That they use the same controller or here - the same bridge and each USB port has it's own power feeding from the built-in PSU. >>>> but that doesn't >>>> change anything since the transferred signal still is routed through the >>>> cabling, which the FW signal isn't. >>> firewire signals are not routed through cables?? so the cables are just >>> for decoration?? >> When you use the term 'routed' you just use it as a 'signal carrier'. >> But a routed signal is depending on other signals in the same chain. - A >> Firewire signal isnot routed - only _carried_ through the cables. > > link? Google is your friend - I guess there are about 100.000 links about Firewire... >>>> FW here is more like a cabled 'radio >>>> transmitter', where you send multiple signals with slightly shifted >>>> carrying signal frequencies to/from each FW device. >>> this just gets even more bizarre. where did you read *that* ? >> Somewhere, > > 'somewhere' ? > > provide a link. this ought to be good. Google is your friend - I guess there are about 100.000 links about Firewire... >> but also tech guys use these terms to explain the way >> Firewire is handled to people, who doesn't know much about computer >> technology. > > that must be why they explained it to you that way. The explanation wasn't to me, but to a totally technically analphabetically person. - I've been using both USB and Firewire since 1997/98, but have now only my scanner, floppydisk drive and a LS-120 drive left using USB. The rest is now either FW400 or FW800 - including professional LaCie music burner. And I can guarantee you that if it wasn't really expensive to exchange the scanner to a FW scanner, then it would have been done long ago. - But a 42bits prof HP scanner with FW is too expensive for me now that I've retired from the computer business because of physically problems with my hips and one shoulder, so the income isn't as high anymore as it has been..... Cheers, Erik Richard -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Erik Richard Sørensen, Member of ADC, <mac-manNOSP(a)Mstofanet.dk> NisusWriter - The Future In Multilingual Text Processing - www.nisus.com OpenOffice.org - The Modern Productivity Solution - www.openoffice.org ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: nospam on 25 Jan 2010 21:40 In article <4b5e4de4$0$8551$ba624c82(a)nntp06.dk.telia.net>, Erik Richard S�rensen <NOSPAM(a)NOSPAM.dk> wrote: > >> Sure it is. And sure it matters how many UsB devices you connect. It is > >> obviously that you've never tried running with more than 1-2 maybe three > >> USB external devices at a time. > > > > i have 7 usb devices on my desktop mac. > > Connected how? I guess not individually - rather by hubs... actually, i miscounted. there are 4 ports on the mac, one of which feeds a 4 port hub, for a total of 7 ports, all of which are in use. however, i forgot about the mouse and printer which are plugged into the keyboard, for a total of 9 devices using two hubs. i have a 7 port hub to replace the 4 port hub, so i have a couple of spare ports, but i've managed without, at least so far. > And you > don�'t have any slow down in speed when transferring to/from > multi-disks? nope. > - If you claim that it is a veeeery special Mac you're > using. I can guarantee you one thing and that /is/ that the speed is > slowed down. I have only 3 USB devices connected, keyboard, trackball > and scanner. And it's for sure that the speed is slowed down when the > scanner is working at the same time as typing in a document! scanning slows down when you type? > > from the developer note for the ibook dual usb: > > > > <http://developer.apple.com/legacy/mac/library/documentation/Hardware/De > > veloper_Notes/Macintosh_CPUs-G3/ibook08May01/ibook-17.html#25515> > > > > The Pangea IC implements two independent USB controllers (root hubs), > > each of which is connected to one of the ports on the back panel of > > the computer. The use of two independent controllers allows both USB > > ports to support high data rate devices at the same time with no > > degradation of their performance. If a user connects a high-speed > > (12 Mbps) device to one port and another high-speed device to the > > other, both devices can operate at their full data rates. > > > > read that last sentence a couple of times. > > Yeah and? - Here's explecitely mentioned _two_USB_controllers_ not _one_ > controller with independant circuits or one controller with dual or > triple circuitry which is the normal on most common PCs. this is about macs, not 'most common pcs.' > But even when there are those two controllers, you will experience some > slow down in speed, since both controllers are relying on the same CPU. > But it's also true that the slow down will be smaller in this construction. what part of "The use of two independent controllers allows both USB ports to support high data rate devices at the same time with no degradation of their performance. " is not clear? note the where it says, 'no degradation of performance'. > > which usb hard drive has both types of ports? > > The French brand, Archos, some of the Mercury enclosures from OWC and > some of the Weiland professional series enclosures. which ones, specifically? i have a few owc enclosures and none have *both* types of usb ports. they generally have a mix of usb, firewire 400, firewire 800 and/or esata, but i've yet to see a hard drive with both a type a & b usb port. the only way that can occur is if it's *also* a hub, and that would be unusual for a basic enclosure. > > i've never seen a drive that can use both usb *and* firewire at the > > same time. some even have physical switches, although most auto-detect. > > which usb drive is this? > > Enclosures from NewerTech. the newertech ministack includes a usb and firewire hub (two separate hubs for two separate connections), as well as a hard drive. is that what you are thinking of? > >> When you use the term 'routed' you just use it as a 'signal carrier'. > >> But a routed signal is depending on other signals in the same chain. - A > >> Firewire signal isnot routed - only _carried_ through the cables. > > > > link? > > Google is your friend - I guess there are about 100.000 links about > Firewire... you made the claim, you provide the link. > >>>> FW here is more like a cabled 'radio > >>>> transmitter', where you send multiple signals with slightly shifted > >>>> carrying signal frequencies to/from each FW device. > >>> this just gets even more bizarre. where did you read *that* ? > >> Somewhere, > > > > 'somewhere' ? > > > > provide a link. this ought to be good. > > Google is your friend - I guess there are about 100.000 links about > Firewire... in other words, it doesn't exist. no surprise there.
From: Daniel Cohen on 26 Jan 2010 07:43
Nick Naym <nicknaym@_remove_this_gmail.com.invalid> wrote: > > Seagate has had a slew of problems these past few years, as evidenced by > posts in its forums as well as user reports on retailer sites like Newegg. > Though not alone in this regard -- overall, drive quality has, IMO, > plummeted because of Industry cost-cutting measures that have resulted in > increasing reliance on economically distressed offshore suppliers -- I have > the sense that Seagate has suffered the most. (As you know, JR, the company > recently slashed its 5-year product warranty to 3 years, using the > transparently lame excuse that it was doing so simply to be "more in line > with the rest of the industry.") I have a Seagate FreeAgent Desktop drive. It goes to sleep after its own period of inactivity (NOT controlled by the Energy Saver settings) and can't be woken again simply. Seagate have a downloadable piece of software that is supposed to cure this, but I am almost certain it is wrongly coded and does not do what it is supposed to. I have mentioned this on the Seagate forums and got no response, although Seagate people do monitor the forums. There doesn't seem to be a way of asking Seagate directly about this. Fortunately I found a way round the issue for my purposes. > Additionally, when I got my Barracuda (September, 2008), I was told by a > Seagate tech that they won't deal with any problems that occur if the drive > is used in an external enclosure (which, of course, is how I'm using it) -- > they don't want to get caught up in a finger-pointing contest between > themselves and the enclosure vendor should a drive problem arise. In another discussion of hard drives on one of the Usenet groups, I commented that this potential issue was a major reason why i bought ready-built external drives (I have one in my own enclosure, that has been fine). -- <http://www.decohen.com> Send e-mail to the Reply-To address. Mail to the From address is never read. |