Slow mac

[robo44]

If i restore my mac to factory settings can i use time machine to set it up and its been playing up lately.
When it starts up in the morning its so slow and sometime it won't log in so i have to re boot it.
I have followed tips online but still its a pain sometimes so i thought about the restore idea.
Any ideas would be appreciated.

 

[maflynn]

If i restore my mac to factory settings can i use time machine to set it up and its been playing up lately.

Restoring to Factory settings, means you don't restore a backup, because what ever is on your computer now slowing it down, will be restored back there again.

What model iMac do you have? What software are you running? Have you run any hardware diagnostics, if you are having trouble logging in, or its very slow, it very well could be a hard drive failing (provided your iMac has a hard drive)

 

[robo44]

Restoring to Factory settings, means you don't restore a backup, because what ever is on your computer now slowing it down, will be restored back there again.

What model iMac do you have? What software are you running? Have you run any hardware diagnostics, if you are having trouble logging in, or its very slow, it very well could be a hard drive failing (provided your iMac has a hard drive)

Ive an 21.5 in iMac Late 2012 running El Captain. How do i run hardware diagnostics ?

 

[keysofanxiety]

Ive an 21.5 in iMac Late 2012 running El Captain. How do i run hardware diagnostics ?

Download SMART Utility here and run it. Anything other than 'Passed' means the hard-drive is failing.

If that comes back clean, open Disk Utility and run a verification on Macintosh HD. If that fails, quote this and we'll run a repair through the recovery partition.

If Disk Utility comes back without an issue (volume appears to be OK in green writing), then try shut down your Mac with the option 'reopen windows when logging back in' disabled (unticked).

Please let me know how you get on.

 

[robo44]

Download SMART Utility here and run it. Anything other than 'Passed' means the hard-drive is failing.

If that comes back clean, open Disk Utility and run a verification on Macintosh HD. If that fails, quote this and we'll run a repair through the recovery partition.

If Disk Utility comes back without an issue (volume appears to be OK in green writing), then try shut down your Mac with the option 'reopen windows when logging back in' disabled (unticked).

Please let me know how you get on.

This is the result

 

[keysofanxiety]

This is the result

Would you be happy to screenshot the outcome instead? Cmd+Shift+4, and click the SMART Utility window.

If that shows 'passed' in green, could you try the other troubleshooting steps?

 

[maflynn]

This is the result

The overall status is showing failing, so I'd say your drive is the issue. Back up your data and get the drive replaced (by yourself or apple authorized dealer).

 

[redheeler]

This is the result

I recommend replacing your mechanical hard drive with an SSD. You can add one internally (a somewhat difficult project to do yourself on that iMac model), or externally using USB 3 or Thunderbolt. A clean install would also be a good idea, and you should always have a backup you can restore from, especially with a hard drive possibly on the verge of failing.

 

[WorkerBee2015]

This is one of the reasons people don't like SMART. It comes up with a change, such as the reallocated sector count, and too many of the utilities erroneously tell people to dump their drive. They swap out their drive and put the "bad" drive into an enclosure, where it lasts 10 more years. Most of the SMART utilities are stealing smartctl from the smartmontools kit, building an interface to it, then claiming it's their own work. You'll find smartctl neatly tucked away inside the package contents, usually in the Resources folder. You're better off downloading smartctl directly and learning how to use it. If someone can't figure out how to program using the IOKit to obtain and analyze SMART data, what makes you think they know enough about SMART to be reporting results correctly? smartctl gives full reports. If the data it produces is too complex, you'd probably be better off just opening up disk utility, clicking on the primary drive, and getting the pass/fail status of the drive there by clicking on the Info button. You can get smartmontools here:

https://sourceforge.net/projects/smartmontools

With that said, is the drive failing? I see no evidence of that. It's not uncommon for a hard drive to successfully reallocate a few questionable sectors, and it's totally normal for an SSD to start doing it as it ages. The normalized value for sector re-allocation is no where near as low as it could be and has not gone down anywhere near the threshold value. (FYI, SMART stats indicate problems when they start going down when the normalized value starts approaching the threshold, which may be the exact opposite of what their raw value indicates, which many find confusing) However, failing/re-allocated sectors in some cases, especially if they start increasing for no apparent reason, can indicate the drive is having problems. I'd be more concerned about failing sectors increasing than a few re-allocated sectors, which might very well have been detected as defective or marginal sectors after the drive was put into use. That doesn't mean you're drive isn't having problems, SMART data collection on hard drives can be horribly erroneous often failing to report some problems.

I'd recommend the following:

• Boot into safe mode and see if the problems continue. If they don't, you likely have a kext extension or startup application that's causing problems.
• Do an SMC reset: https://support.apple.com/en-us/HT201295
• Do an NVRAM reset: https://support.apple.com/en-us/HT204063
• Try a different OS.
• Try a diagnostics app like Apple Hardware test, Scannerz, or TechTool pro. They can all find different problems but the latter 2 cost money.

With that said, if you went with an SSD or one of the newer AF formatted drives you would definitely see a performance improvement. Although the newer AF formatted drives with high areal density aren't as fast as SSDs they're typically much faster than Apple's original OEM HDDs, and some people still need pretty big drives, making SSDs economically unfeasible.

 

[ApfelKuchen]

On an older machine you can often have issues related to old/outdated startup items (both in System Preferences > Users & Groups > Login Items and also inside the system library). I suggest running etrecheck (donation ware - etrecheck.com) and see what it finds.

 

[keysofanxiety]

This is one of the reasons people don't like SMART. It comes up with a change, such as the reallocated sector count, and too many of the utilities erroneously tell people to dump their drive. They swap out their drive and put the "bad" drive into an enclosure, where it lasts 10 more years. Most of the SMART utilities are stealing smartctl from the smartmontools kit, building an interface to it, then claiming it's their own work. You'll find smartctl neatly tucked away inside the package contents, usually in the Resources folder. You're better off downloading smartctl directly and learning how to use it. If someone can't figure out how to program using the IOKit to obtain and analyze SMART data, what makes you think they know enough about SMART to be reporting results correctly? smartctl gives full reports. If the data it produces is too complex, you'd probably be better off just opening up disk utility, clicking on the primary drive, and getting the pass/fail status of the drive there by clicking on the Info button. You can get smartmontools here:

https://sourceforge.net/projects/smartmontools

With that said, is the drive failing? I see no evidence of that. It's not uncommon for a hard drive to successfully reallocate a few questionable sectors, and it's totally normal for an SSD to start doing it as it ages. The normalized value for sector re-allocation is no where near as low as it could be and has not gone down anywhere near the threshold value. (FYI, SMART stats indicate problems when they start going down when the normalized value starts approaching the threshold, which may be the exact opposite of what their raw value indicates, which many find confusing) However, failing/re-allocated sectors in some cases, especially if they start increasing for no apparent reason, can indicate the drive is having problems. I'd be more concerned about failing sectors increasing than a few re-allocated sectors, which might very well have been detected as defective or marginal sectors after the drive was put into use. That doesn't mean you're drive isn't having problems, SMART data collection on hard drives can be horribly erroneous often failing to report some problems.

I'd recommend the following:

• Boot into safe mode and see if the problems continue. If they don't, you likely have a kext extension or startup application that's causing problems.
• Do an SMC reset: https://support.apple.com/en-us/HT201295
• Do an NVRAM reset: https://support.apple.com/en-us/HT204063
• Try a different OS.
• Try a diagnostics app like Apple Hardware test, Scannerz, or TechTool pro. They can all find different problems but the latter 2 cost money.

With that said, if you went with an SSD or one of the newer AF formatted drives you would definitely see a performance improvement. Although the newer AF formatted drives with high areal density aren't as fast as SSDs they're typically much faster than Apple's original OEM HDDs, and some people still need pretty big drives, making SSDs economically unfeasible.

I'm sorry to call you out so passionately, but what rubbish this is. A drive will often still work with reallocated sectors. That doesn't mean you should keep using it. The performance will be vastly reduced. Hard-drives are massively unreliable due to how the technology works.

If there's a SMART error, replace the drive. It's as simple as that. Absolutely nobody in this industry would recommend continuing using a SMART error/fail drive as your main drive. As an external, maybe. But your main drive? No way.

And for you to say that it could be a kext extension or startup application that's causing the problems, after the SMART has failed, is ludicrous misinformation.

 

[Weaselboy]

Most of the SMART utilities are stealing smartctl from the smartmontools kit, building an interface to it, then claiming it's their own work.

Is there a particular utility that is doing this that you are concerned about? There are some GUI SMART utilities that use smartmontools and there is absolutely nothing wrong with that. smartmontools and smarctl is open source and the license allows its use in other utilities as long as credit is given. For example I use DriveDX and if you look in the acknowledgements menu under help you will see the required smartmontools disclosure.

This is a very common usage of open source software and is certainly not stealing anything. The popular cloning app Carbon Copy Cloner includes the open source rsync binary in the same way. Even OS X includes quite a bit of open source software.

 

[WorkerBee2015]

I'm sorry to call you out so passionately, but what rubbish this is. A drive will often still work with reallocated sectors. That doesn't mean you should keep using it. The performance will be vastly reduced. Hard-drives are massively unreliable due to how the technology works.

If there's a SMART error, replace the drive. It's as simple as that. Absolutely nobody in this industry would recommend continuing using a SMART error/fail drive as your main drive. As an external, maybe. But your main drive? No way.

And for you to say that it could be a kext extension or startup application that's causing the problems, after the SMART has failed, is ludicrous misinformation.

I suggest you learn more about SMART before speaking. What I don’t like about the results in the picture is quite simple: The SMART output does not clearly show that the drive is failing. The application detected a change but the drive itself is not registering a failed status. The application itself decided that this drive is failing, not the internal SMART monitoring software.

SMART comes with a Normalized Value, a Worst Value, a Threshold Value, and a Raw Value. Depending on how the internal SMART software stores data, the Raw Value may or may not be understandable to an end user. In this case it’s likely a safe bet the raw value is indicating the number of sectors reallocated. As a unit with SMART starts having problems, the Worst Value may start dropping toward the threshold, and when it actually meets or falls below the threshold, the “Failed” status will go from “Never” to, depending on the software, a date and time, possibly the word “Failed,” and for some software if the failures are beginning to occur more frequently and they begin to approach the threshold, it may say “Failing”. I’m speaking generically about SMART monitoring software. Items which can indicate possible failing conditions will be designated as “Pre-fail” while the others may change as the drive ages or, especially for SSDs, be informational (like power on hours) and are often of the type “Old age”. Many people find this confusing because in some cases, as the raw value may actually be going up.

If you look at the data, the SMART normalized and worst values are 100, the threshold is 5, and the failed status is “Never.” It is not abnormal for a drive to have a few bad sectors on it and to successfully re-locate them. It is totally normal for an SSD to start developing bad NAND blocks as it ages.

What a user needs to be aware of is whether or not these values are changing frequently. A one time event, such as a drive detecting a few sectors that were bad when the drive left the factory, or for that matter even a one time minor to moderate head crash that was correctable does not mean a drive is failing, it means just what it says: it found some bad sectors so it replaced them. If values are changing over time it indicates a problem, and probably a serious one. If so many sectors needed to be replaced that the supply of spares is exhausted, normally the reallocated sector count worst value would drop down to or below the threshold and the status would change to either a failed status or a failing status.

Another problem with SMART is that too often it simply doesn’t register any problems when problems are obvious. The fact is the engineers designing this cannot account for every single possible failure mode possible. All they can do is develop what they see as the most likely failure events and hopefully the SMART software will pick it up.

Does this mean the drive doesn’t have problems? No, and I never said that. There may be problems developing. My problem with the application is that it’s indicating a failed status when one doesn’t exit. Instead, it should be informing the user that a change occurred on a Pre-fail item. If all this is was the drive detecting a few bad sectors that didn’t get picked up during manufacturing, he’ll be throwing a good drive away and he’ll still be stuck with his original problem.

 

[keysofanxiety]

I suggest you learn more about SMART before speaking. What I don’t like about the results in the picture is quite simple: The SMART output does not clearly show that the drive is failing. The application detected a change but the drive itself is not registering a failed status. The application itself decided that this drive is failing, not the internal SMART monitoring software.

SMART comes with a Normalized Value, a Worst Value, a Threshold Value, and a Raw Value. Depending on how the internal SMART software stores data, the Raw Value may or may not be understandable to an end user. In this case it’s likely a safe bet the raw value is indicating the number of sectors reallocated. As a unit with SMART starts having problems, the Worst Value may start dropping toward the threshold, and when it actually meets or falls below the threshold, the “Failed” status will go from “Never” to, depending on the software, a date and time, possibly the word “Failed,” and for some software if the failures are beginning to occur more frequently and they begin to approach the threshold, it may say “Failing”. I’m speaking generically about SMART monitoring software. Items which can indicate possible failing conditions will be designated as “Pre-fail” while the others may change as the drive ages or, especially for SSDs, be informational (like power on hours) and are often of the type “Old age”. Many people find this confusing because in some cases, as the raw value may actually be going up.

If you look at the data, the SMART normalized and worst values are 100, the threshold is 5, and the failed status is “Never.” It is not abnormal for a drive to have a few bad sectors on it and to successfully re-locate them. It is totally normal for an SSD to start developing bad NAND blocks as it ages.

What a user needs to be aware of is whether or not these values are changing frequently. A one time event, such as a drive detecting a few sectors that were bad when the drive left the factory, or for that matter even a one time minor to moderate head crash that was correctable does not mean a drive is failing, it means just what it says: it found some bad sectors so it replaced them. If values are changing over time it indicates a problem, and probably a serious one. If so many sectors needed to be replaced that the supply of spares is exhausted, normally the reallocated sector count worst value would drop down to or below the threshold and the status would change to either a failed status or a failing status.

Another problem with SMART is that too often it simply doesn’t register any problems when problems are obvious. The fact is the engineers designing this cannot account for every single possible failure mode possible. All they can do is develop what they see as the most likely failure events and hopefully the SMART software will pick it up.

Does this mean the drive doesn’t have problems? No, and I never said that. There may be problems developing. My problem with the application is that it’s indicating a failed status when one doesn’t exit. Instead, it should be informing the user that a change occurred on a Pre-fail item. If all this is was the drive detecting a few bad sectors that didn’t get picked up during manufacturing, he’ll be throwing a good drive away and he’ll still be stuck with his original problem.

I fear this is another example of somebody whose knowledge is blinding them to the obvious answer.

OP mentions system is running slowly. SMART shows reallocated sectors. Anybody worth their salt would put the two together.

Instead you bleat about SMART reliability and how an HDD can still be usable despite this and that — and then go onto suggest that it's a different problem causing the issue they're having. I can't believe what I'm reading.

Furthermore, OEMs replace HDDs and SSDs without question if SMART has a warning — even ONE reallocated sector — well, at least there's been no issue doing that with a Toshiba/Fujitsu/Apple repair centre.

With HDDs especially, and to a point, technology in general, it's always better to be safe than sorry. Any amount of experience in the field should surely tell you that. And if you have the arrogance to not err on the side of caution as you think you know a little better, just because SMART status can occasionally cause a few hiccups, or misread, or not hit the 'definite failure' threshold, or not be completely accurate, then frankly that says more about your ego than your knowledge.

 

[WorkerBee2015]

Is there a particular utility that is doing this that you are concerned about? There are some GUI SMART utilities that use smartmontools and there is absolutely nothing wrong with that. smartmontools and smarctl is open source and the license allows its use in other utilities as long as credit is given. For example I use DriveDX and if you look in the acknowledgements menu under help you will see the required smartmontools disclosure.

View attachment 628804

This is a very common usage of open source software and is certainly not stealing anything. The popular cloning app Carbon Copy Cloner includes the open source rsync binary in the same way. Even OS X includes quite a bit of open source software.

Linux has a version of an interface for smartctl and they openly advertise it as such. Some of the companies that are using smartctl for their SMART monitoring either conceal or bury the fact that smartctl is doing the work for their product, and I assume it's because they don't want users to know that what they're paying for can be had for free. There are problems with this:

smartctl has a database that gets updated with new drives and firmware updates. It also has a set of "default" parameters it loads. Different drives may use some of the default parameters or they may override them. For example attribute 9 may have a default of power on hours, but on another drive or SSD, the default may be replaced by a different parameter, and possibly a pre-fail type. These can change with firmware updates. If the user is unaware of the smartctl database changes, the output may erroneously map a default value where a custom value needs to be in place. If you download the code to smartmontools and look at the file named "drivedb.h" you will see all the overrides and additions.

If someone is unaware of the fact that smartctl is doing the actual work and they apply a firmware update and that value is changed or another is added, then eventually that change will be added to the smartctl database and the user will know to download and install it - and that's the problem. If the user doesn't know to download and install a new version of smartctl, then what he or she may be seeing is erroneous output.

The drivedb.h file is here:

https://sourceforge.net/p/smartmontools/code/HEAD/tree/trunk/smartmontools/drivedb.h

For example, if you look at the screenshot above, attribute 160 says "Unknown Attribute." Is it, or is that version of smartctl outdated to the point that the info is invalid? If this info is possibly invalid or it's using default parameters when it shouldn't be, is it not possible it's reporting completely erroneous information to a user?

 

[Weaselboy]

Linux has a version of an interface for smartctl and they openly advertise it as such. Some of the companies that are using smartctl for their SMART monitoring either conceal or bury the fact that smartctl is doing the work for their product, and I assume it's because they don't want users to know that what they're paying for can be had for free.

You did not answer my question. Specifically which OS X SMART GUI utility is it you think is "stealing" smartctl as you said. The one I use I showed exactly where they are disclosing the use of smartctl as required by the open source license. There is absolutely nothing wrong with this.

So when smartctl posts an update due to new drives/firmware, then the dev. who made the GUI utility incorporates the smartctl changes into a new version and releases it. In the time I have been using DriveDX this has happened a couple times.

If you prefer to use the command line smartctl utility, I totally understand that. But there is nothing wrong at all with using a GUI interface for that utility and it is in no way stealing like you said.

 

[ApfelKuchen]

Yeah, "Hard drives are the root of all evil." Only, not necessarily. There are many potential causes for a slow Mac (my favorite culprit tends to be bad startup/login items, and near-full disk is another). Sure, SMART shows 4 reallocated sectors. 0 errors, 0 CRC errors. Every entry in the "Failed" column is, "Never." That's hardly the end of the world. There are roughly 2 million sectors per GB of storage (at 512 bytes per sector), so since we're talking about a 1TB HDD... four out of 2 billion sectors? Care to calculate the percentage?

Sure, if I was running a data center, I'd probably be all over SMART status, and be pro-actively replacing hot-swappable drives right and left. But a hard drive replacement in a 3-4 year old IMac? Not the first thing I'd recommend due to its cost and complexity, especially if a restore from Time Machine backup afterwards might (or might not) resurrect some of the original problems.

Here's the thing about status-monitoring software. Judgements like "failing" are just that - judgements. What criteria go into those judgements? What is the target market of the product? What is "Aging" and "Failing?" https://www.volitans-software.com/support/smart-utility-faq/ - having read that FAQ...

In the test results the OP shared, all those "aging" and "failing" labels in the Type column are accompanied by a "Never" in the "Failed" column. Yet despite all those "Nevers," the overall result is "Failing?" Until there is a value other than "Never" in "Failed," then, by the developer's own definition, that particular item is neither aging nor failed.

Yeah, the developer believes that having just a few bad sectors means the drive is failing. Yet he says, "Some people had drives that with a few bad sectors, that lasted for years but in my experience, when a couple of bad sectors are found, more will follow." Yes, more will likely follow. That's the way the world works. I've been having a few "senior moments" lately. It must be time for assisted suicide.

Again, he says, "In the end, it is up to you whether to replace the drive or to take a chance with your data on that drive, hoping that it will last for years." Seems pretty dire, except this is exactly why we make backups. If your customer base is made up of people who are worried about failing hard drives, you're probably better off playing to those fears than saying, "Nothing to worry about, just make sure you have a good backup!"

I've yet to replace the HDD in my Early 2008 iMac. Been running betas on that thing for years, had plenty of slowness issues - did an erase/reinstall OS X/restore from Time Machine a few months back - big help. At other times, Etrecheck found a number of software-based issues that also made a huge difference. Altogether, the thing is performing better in 2016 than it did in 2013 when it was demoted to "Mac #2."

I'm not saying my issues and solutions will necessarily be the next person's issues and solutions. However, these are complex systems - simplistic approaches like, "When in doubt, blame the HDD" just don't cut it.

 

[killerovsky]

Most of the SMART utilities are stealing smartctl from the smartmontools kit, building an interface to it, then claiming it's their own work.

Linux has a version of an interface for smartctl and they openly advertise it as such. Some of the companies that are using smartctl for their SMART monitoring either conceal or bury the fact that smartctl is doing the work for their product, and I assume it's because they don't want users to know that what they're paying for can be had for free.

There are dozens of commercial apps that are somehow using smartmontools kit, but one link can tell a lot more than a thousand words - https://www.smartmontools.org/wiki/Links#Graphicaluserinterfacesforsmartctl

 

[cynics]

Raw value of reallocated sectors is 4, current pending sector is 0.

Look elsewhere for your OS X slow down problems.

Try signing in as a guest user and see if there is a performance increase.