Q: Replace Hard Drive with SSD
-
May 4, 2015 11:14 AMby MrWilliams201,Hello:
I guess now I need to do the clarification.
1. The Hitachi is fast for a hard drive, especially when compared to other hard drives
2. Partitioning out on an HD is different than it is on an SSD and an HDD
For item 1, it's self explaining but theoretical highest media transfer rate is 171MB/sec:
http://www.storagereview.com/hitachi_travelstar_z7k500_review
For item 2, on an SSD the physical media is "moving" meaning it's not locked to specific chips. i.e. if you go to block XXXX on an HDD, it will be physically located at a specific location on the drive. If you do the same with an SSD, as files are written to, erased and re-written to the physical location of the chip being used to store specific bytes is not XXXX, it might be XXXX the first time, YYYY the second, and ZZZZ the third.
What was being referred to as partitioning out the data to stop access to a specific region of an HDD with damage, I don't think would work on an SSD. Maybe it can and I don't know any better, but I don't think that would work. In fact I would think that if a bank of chips on an SSD were bad, I would think the SSD would be pretty much unusable.
I may be wrong, but I don't think so.
-
May 5, 2015 11:00 AMby MrWilliams201,That's true, but in an SSD it's not the same thing. On a hard drive he would be partitioning out specific regions of the physical media (those that are damaged) but on an SSD the partition is essentially virtual and there's no guarantee the contents could be tied to specific locations on the SSD.
This is the second time I've replied to this. When I did this the first time my previous post didn't show up and I assumed I simply didn't save it properly.
Now, can someone explain that to me? It's like the post didn't load when I loaded the thread.
-
May 6, 2015 11:20 AMby HuntsMan75,I appreciate the repartitioning trick but my system isn't some old computer that's 10+ years old and out of date. With new hard drives costing tens of dollars if something goes wrong with one IMHO you might as well replace it. Even the smaller SSDs are beginning to get cheap.
-
May 7, 2015 10:47 AMby MrWilliams201,I don't think anyone was implying that partitioning out a section of a drive to keep one with problems usable was the intention, but rather a description of what someone could do with an old system that's not worth spending money on and it may even be difficult to get parts for it.
-
May 8, 2015 11:04 AMby ThomasB2010,It was never intended to be a viable or reliable solution to a problem. I did this on an old clunker because it was getting minimal use, I didn't want to open it up, I didn't want to spend money on it, and I don't even know if IDE drives that work with it anymore are even available at a reasonable price. It might come in handy in some type of emergency, and I suppose it might, as was my case, continue running for a long time without any further problems, but if this was my primary system, the drive would have been replaced.
-
May 9, 2015 11:47 AMby HuntsMan75,From this previously linked article:
http://www.tweaktown.com/news/41180/third-party-ssd-vendors-address-apple-trim-i ssue/index.html
I get the quote:
"Apple recently upset their users by removing support for third-party software that enables TRIM functionality. Perhaps most distressing was the fact the change went unannounced. Many Apple users with Trim Enabler, a third party app that enables TRIM functionality, unfortunately bricked their installs when they updated to OS X 10.10 Yosemite. There is a method for enabling TRIM with third-party SSDs, but it involves creating a massive security vulnerability."
Just exactly how "massive" is this security threat? For this to be a threat wouldn't someone already need access to your system, as is in be on it, as in it's already compromised? Was this thread present in previous MacOS versions or it this a Yosemite threat only.
-
May 9, 2015 1:05 PMby FatMac>MacPro,HuntsMan75 wrote:
...Just exactly how "massive" is this security threat? For this to be a threat wouldn't someone already need access to your system, as is in be on it, as in it's already compromised? Was this thread present in previous MacOS versions or it this a Yosemite threat only...The security threat (that unsigned kexts can be installed on your system), has been around right through Mavericks. They can be installed or modified (as Trim Enabler does) as part of an application installation. The degree of risk is very much a matter of your level of care in choosing the source of the software you're installing.
The "method for enabling TRIM with third-party SSDs" is the turning off of the check for proper signing that was built into Yosemite. With that method applied to a Yosemite installation by TRIM Enabler, Yosemite will continue to work. But if, for any reason, that signing check is turned back on (Cindori explains what might do that), the Mac won't boot. The link provides a series of steps to undo the booting failure once it occurs, though the primary advice is to turn TRIM off before any PRAM reset or hardware maintenance to avoid the problem in the first place.
Frankly, the best way to avoid the boot problem is to stay with Mountain Lion or Mavericks and continue to be careful with what third party software you install.
-
May 9, 2015 4:27 PMby HuntsMan75,Well I guess I can sort of understand their logic. My sister stumbled on some web site that informed her there was something wrong with her system (I think it suggested it had a virus) so she then proceeded to download MacKeeper. When I heard about this I asked her how she figured a website was able to figure out her system had a virus, then I briefly explained to her what had just happened, then I proceeded to tell her step by step how to get rid of it. I'm sure there are probably others out there that would do the same thing, only with much more malice.
However, I also thought that kexts could be signed using a developer account and registering it with Apple. That's what I though anyway. I know a regular application can do that but a fair number of developers don't do it.
-
May 9, 2015 8:01 PMby FatMac>MacPro,HuntsMan75 wrote:
...However, I also thought that kexts could be signed using a developer account and registering it with Apple. That's what I though anyway. I know a regular application can do that but a fair number of developers don't do it.I think you are correct. But according to Cindori, "Have you ever wondered why disk manufacturers like Corsair, Samsung etc. have never written a Trim driver for OS X?...The reason is that Apple is keeping the AHCI driver private, making it impossible for anyone, even these big companies, to create a driver for SSD’s. So the issue is not that Apple “doesn’t allow Trim Enabler”, but that they don’t allow anyone to create AHCI drivers." So signing the kext would be the easy part. The current AHCI driver for TRIM works very well with third party SSDs but only after its test for Apple branding of the installed SSD is disabled, which is what TRIM Enabler does; that breaks the kext's signing, which Yosemite checks for, and if the signing is broken, when Yosemite's signing check is next turned on, that breaks Yosemite.
Checking for kext signing is a good thing, but so is having TRIM functional, so it's a choice that had to be made with the advent of Yosemite (and probably subsequent versions of OS X). The difference is that careful and knowledgeable installation of software can, and already has reduced the likelihood of installing things like MacKeeper (I've gotten phone calls telling me my Windows computer has a virus ["how come you can tell from a distance I have a virus but you can't tell that I have OS X?"]), but nothing the user can do will provide the SSD controller what TRIM does except turning the kext signing test off and TRIM on, via TRIM Enabler.
The risk there is breaking Yosemite, and Cindori's instructions on how to back out of that if it happens work but need a level of sophistication to implement that I suspect an experienced Mac user can manage but perhaps not someone drawn to the Mac by his or her iPhone experience. (Locally, there has been an upsurge in car break-ins and thefts and the solution the authorities have is to put placards on windshields to remind drivers not to leave ignition keys or money visible in their cars - what's the equivalent in cautious computing?) Indeed, I think the best way to be prepared for that eventuality is to create a bootable Yosemite installation on a Flash drive, then copy the Terminal commands in Cindori's instructions to a TextEdit document, substituting the Volume name of your boot drive in each command. Then, recognizing and remembering the risk, for any OS update (e.g., 10.10.1 to 10.10.2) or test or fix that changes the PRAM, use TRIM Enabler to turn TRIM off and the kext signing test back on first. If it breaks anyway, boot from the Yosemite flash drive you prepared and follow the customized instructions you also prepared to get running again.
It's a pain, as is locking your car and taking your keys, but less so than having your car stolen or your SSD working harder than it needs to.
-
May 10, 2015 6:00 PMby ThomasB2010,Isn't the NVRAM susceptible to corruption, hence articles like this:
How to Reset NVRAM on your Mac - Apple Support
Would I be safe in assuming that if they became corrupt, every time the NVRAM is reset I'd have to re-do the Cindori steps?
-
May 11, 2015 8:41 AMby FatMac>MacPro,ThomasB2010 wrote:
...Would I be safe in assuming that if they became corrupt, every time the NVRAM is reset I'd have to re-do the Cindori steps?If you reset the NVRAM before you turned off TRIM using TRIM Enabler to revert to the pre-TRIM enabled condition, you would indeed have to re-do those steps, which have some wrinkles I discovered when trying it myself. However, the NVRAM can be corrupted without the Mac being disabled, so you'd have the opportunity to cleanly disable TRIM temporarily before reset. The user would just have to remember to take that opportunity first, and there's the rub.
