If abnormal noises are confirmed to be coming from the drive, you should back up its contents as soon as possible. Common noises coming from a mechanical hard drive experiencing problems can include the following: • Excessively loud clicking noises • A repetitive "chug-a-chug" sound • Screeching sounds • A very loud humming sound • Grinding noises Most of the noises above are a clear indicator of a drive that's about to fail, or beginning the initial stages of catastrophic failure. Other noises, such as the "chug-a-chug" sound usually indicate a bad or weak sector, which doesn't necessarily mean the drive is about to fail, but has surface platter damage. We recommend becoming familiar with the noises coming from a mechanical hard drive when it's new so you'll be able to distinguish the normal from the abnormal. It is not uncommon for a drive to become louder as it ages. All drives make noise, but newer drives are nearly silent. If you listen to a drive in good working order operating normally it may make a faint noise, similar to the sound of rattling, with intermittent light clicking noises, with both being somewhat random in nature. Loud clicking noises are usually indicative of catastrophic failure, particularly if they're repetitive. This usually indicates that the drive heads are slamming against the extremes of the drive mechanism. In most cases, this is a catastrophic failure of the hard drive. This type of noise most often comes on suddenly, and it will be much louder than the light clicking noises you may hear from a newer drive in good working order. Screeches and grinding noises, likewise, are usually signs of impending catastrophic failure. These types of noises may initially start appearing in short bursts that become increasingly more common as time progresses, until the drive eventually totally fails. The causes can be numerous, with none of them good, and include failing bearings, failing motors or actuators, or in some cases the dragging of the drive heads across the surface of a hard drive platter. If you have a drive in this type of condition, we recommend getting data off the drive as soon as possible. The drive may fail in a matter of minutes, hours, or even days, but in nearly all cases, failure is imminent. The "chug-a-chug" noise is usually caused on a mechanical hard drive by a bad or weak sector. On some contemporary hard drives, which are extremely quiet compared to drives ten or more years older, this noise may not be immediately evident. Delays will almost always accompany these, possibly with the appearance of a "spinning beach ball." The noise itself, if audible, is caused in mechanical hard drives by the drive continually repositioning itself at the start of a bad or weak sector, attempting to read it, and when it can't be read, it repositions the drive heads to the start of the bad sector and starts over again. A bad or failed sector will never succeed in this operation, whereas a weak sector will eventually be able to be read. More information is provided below. Delays from bad or weak sectors can occur in a mechanical hard drive, and bad blocks (sectors) can occur on SSDs. Because the two behave differently, we will treat both types of storage separately. On a mechanical hard drive, if a sector is bad, it means the sector cannot be read at all, and after a certain number of retries, the hard drive controller will give up and issue an I/O error. This usually takes seconds and the delay is noticeable. Usually the operating system or the application will give an indication that the read or write operation failed. In some cases, the problem may be corrected automatically by the drive controller, but data in such a sector cannot be recovered, and the file associated with that data will likely be corrupt and may not be recoverable at all. Bad sectors can exist in regions of a drive that are not used and unless the drive is being tested using a tool designed to detect these types of problems, will remain in place until the drive attempts to use the bad sectors. Increasing numbers of bad sectors, even in unused regions of a drive, can be a sign of increasing problems. A weak sector is a sector that can be read, but only after the drive controller reads it many times. This can cause delays on the order of milliseconds to tens of seconds. The drive controller will attempt to read the sector and may average the data recovered until it can verify that what's being read is accurate. Weak sectors will only be obvious during a read operation. Weak sectors can be very problematic if they exist in a file such as a cache file that's continually read, over and over again. This will cause considerable and noticeable delays in system performance. Bad and weak sectors on mechanical hard drives can be caused by a hard drive head crash or degradation of the media on the surface of a hard drive. If a head crash has occurred, the drive heads have made contact with the surface of the platter and damaged it. It is not uncommon for a drive that's experienced a head crash to consist of regions of both bad and weak sectors. The severity of a crash may be limited to a few sectors, or widespread, spanning gigabytes of data. If the media has an area that's degraded, it may be due to a slight manufacturing defect in the surface of the platters or some of the sectors may be beginning to lose their capacity to retain information. A mechanical hard drive can correct bad sectors on a hard drive by remapping the bad sectors to spare sectors. This effectively removes the bad sectors from the available storage blocks in the drive and replaces them with good sectors. This will occur only during a write operation. If bad sectors develop in a pre-existing file, they will remain in place until until the file is deleted or the area is written over during a zeroing operation (a reformat of the drive.) Deleting a file containing damaged sectors will make the damaged sectors available for re-writing, and ideally, the next time the controller attempts to write to those sectors, it will detect them as bad, and then map out. Hard drives have a limited number of spare sectors, and if the supply becomes exhausted, in many cases the controller will just leave the bad sectors in place. Weak sectors are particularly problematic because a hard drive may see them as viable, even though the sectors are causing delays. How a particular drive controller determines which weak sectors qualify as being ready for remapping is completely at the discretion of the manufacturer and will likely vary widely from model to model. As with bad sectors, if a pre-existing file contains weak sectors, no attempt will be made to overwrite or remap the sectors until the file is deleted or the drive is zeroed. SSDs generally cannot develop weak sectors, but they can develop bad blocks - in fact it's normal for them to do so. An SSD is totally electronic and totally digital. Unlike a mechanical hard drive that's converting an analog signal from the surface of a rotating platter to a digital format, an SSD stores it's data in digital format. The media on a hard drive, assuming it's undamaged, can be used to read and write data countless times until the media itself begins to fail to retain information, which is often on the order of a decade or longer. The memory cells on an SSD have a limited write cycle count, meaning that after they've been subjected to a certain number of writes, they fail and need to be replaced by spare blocks, similar to the spare sectors on a mechanical hard drive. As you might guess, the number of spare sectors on a mechanical hard drive is small compared to the number of spare blocks available on an SSD. Ideally, the way an SSD handles block management is by monitoring write counts and data retention capability when writing data. If the write count is high and the ability to retain information appears to be marginal, the block is excluded from use and replaced by a spare block. Failing blocks will be re-mapped until the set of spares is exhausted. According to manufacturers, this typically occurs in terms of tens of years, not years. Unfortunately, some blocks on SSDs simply refuse to behave like they're supposed to. They will simply go bad, for what appears to be no obvious or apparent reason. The problem is similar to a bad sector on a hard drive caused by a media defect. As with mechanical hard drives, the problem is somewhat of a rarity, but when it occurs, it will be identified by the system as a bad block. Like a mechanical hard drive, if this bad block is discovered during a write operation, the controller should replace it with one of the spares, automatically correcting the problem. Unfortunately, if the failure occurs after data has been written to it, the controller will likely make no effort to correct it unless the file containing the bad block is itself modified. When a bad block of this type exists, the results as far as the system goes, will be identical to a bad sector on a hard drive. The system will attempt to continually read the file containing the bad block, it will fail every time, and the system will finally issue an I/O error. Typically the operating system or the application will issue some type of message indicating the operation did not succeed, and with a Mac you may witness the "spinning beach balls" as the system locks up and attempts to read the bad block. Note that this phenomena is a rarity, and if you've recently purchased an SSD and this is a recurring problem, it's advisable to ensure the appropriate drive management software has been installed and that the drive's firmware is up to date. A new SSD experiencing excessive bad blocks is an indication that the memory used in the SSD is likely defective, the support software isn't installed, or the controller is not functioning properly. Periodic delays while using the system may be caused by many things. If the problems are tracked down to a drive using a test tool and a mechanical hard drive is the culprit, the most common cause is that the drive platters have, or are developing bad or weak sectors. If the drive is an SSD, it's quite possible that it's developed one or more bad blocks in such a state that the controller cannot or will not automatically correct them. Bad or weak sectors on a mechanical hard drive and bad blocks on an SSD were previously covered, so in this section we'll look at other causes. Other types of periodic delays can be caused by hardware malfunctions, software problems, and user problems. In most cases, hardware problems will be something fairly mundane, but never the less, not necessarily easy to identify. Software problems can be caused by applications as well as bugs in the operating system. User problems typically occur when the user is unaware that something they're doing is resulting in the problem or they're not familiar with the behavior of one or more system components. Hardware problems are usually either catastrophic or intermittent. A catastrophic problem will usually cause the system to fail to even successfully start, crash during boot time, or if it's associated with a specific component that's not in use all the time, crash when the system attempts to use that resource. These types of problems typically don't cause delays, they cause system crashes. Most periodic delays that are sourced to hardware are caused by intermittency, and if intermittent connections exist, delays may appear. The most common cause of intermittency in a system are the cables and connections between the logic board and an internal hard drive, or in the case of an external drive either a bad cable or a failing connector (or one that's not fully plugged in.) Secondary problems can be caused by something that's usually plugged directly into the logic board, such as an Airport card, a Bluetooth card, etc. If the internal cable between the hard drive and logic board has problems, whether it's an SSD or mechanical hard drive, it will usually cause delays, and may cause the system to generate periodic I/O errors. The symptoms are nearly identical to bad blocks or bad/weak sectors with one critical exception: bad blocks on an SSD or bad/weak sectors on a mechanical hard drive will always occur at the exact same places during a surface scan test, whereas problems caused by intermittent connections will occur at random, independent of a surface scan test. You cannot simply look at log files for the presence of an I/O error and assume there's a bad block or sector in the media, because a bad cable can cause the exact same problems and delays. If the problem cannot be correlated to the drive or cable, it's possible that intermittent connections exist somewhere else in the system. Anything that connects to the system that's in frequent use, such as the keyboard, Airport card, or Bluetooth card (to name a few) may develop problems with both connections and in the cables themselves. Much in the same way that a bad hard drive cable may cause intermittent delays or I/O errors, the same can often happen if anything plugged into the logic board fails to communicate with the logic board properly. Intermittency can cause a peripheral to disconnect and re-connect, causing the system to interrupt itself, re-establish communications with the device, only to have it disconnect again, and repeat this cycle over and over again. Do not assume the logic board needs to be replaced if problems cannot be traced to the hard drive and/or its cables and connections or software problems.
FSE
All Software and Computer Systems Company, LLC logos are a trademark (TM) of Software and Computer Systems Company, LLC. Scannerz, Scannerz Lite, Scannerz with FSE-Lite, Scannerz with FSE, FSE, Phoenix for Mac OS X, Performance Probe for Mac OS X, SpotOff, and Spot-O-Meter are trademarks (TM) of Software and Computer Systems Company, LLC. All software produced and licensed by Software and Computer Systems Company, LLC is copyright© Software and Computer Systems Company, LLC 2005 - 2015. The contents of all pages and images contained in this web site are copyright© Software and Computer Systems Company, LLC, 2005-2015. Apple is a trademark of Apple Inc., registered in the U.S. and other countries. Apple Macintosh, Mac, the Mac Logo, and MacOS are registered trademarks of Apple Inc, in the U.S. and other countries. PowerPC™ is a trademark of International Business Machines Corporation. Intel is a trademark of Intel Corp. in the U.S. and other countries.
General Symptoms of Hard Drive, SSD, System, and Software Problems
Updated March 31, 2015
In this article, we will be describing problems with both traditional hard drives as well as solid state drives (SSDs.) If the problems are applicable to both, then the term "hard drive" or just "drive" will be used. When there are variations between the two, we will identify those specific to SSDs and traditional, electro-mechanical hard drives accordingly. What appears to be a drive problem may be caused by any of the following: • A genuine drive problem • A problem with any of the system components supporting the drive • System problems that have nothing to do with the drive • Operating system bugs or other software related problems • User problems
General Symptoms
The only way to determine the root cause of the problem is by investigating it. Typically, if a hard drive or hard drive related problem is occurring, a user may experience any of the following symptoms: • Abnormal noises (generally not applicable to SSDs) • Delays from bad or weak sectors on mechanical hard drives and bad blocks on SSDs • Periodic delays that weren't present before or have become increasingly more common. • Failure to successfully read or write data to the drive • The file system frequently needs to be "repaired" by Disk Utility • Running applications lock up for no apparent reason • The drive is no longer recognized by the system • A boot drive will no longer boot • A drive ejects itself for no apparent reason or won't mount • A drive appears to shut itself off for no apparent reason Unfortunately, other problems, such as logic board problems, cable problems, software problems, or system oddities may produce similar symptoms. The most common symptom of a drive problem, whether SSD or mechanical hard drive, will be delays and lockups, typically accompanied by "spinning beach balls." Below we will address we will address the topics above in more detail. Abnormal noises are almost always a sign of a failing mechanical hard drive, provided the noises are actually originating from the drive and they're abnormal. SSDs are totally electronic and cannot make noise unless a component, such as a capacitor or other circuit board component is literally in the process of failing catastrophically, which is highly unlikely. Most computers use cooling fans, and if the bearings fail or seize up, the noises they generate may sound nearly identical to those of a failing hard drive. Other devices, such as optical drives may cause noises that my delude a user into believing their hard drive has problems when it doesn't. If abnormal noises can be traced to the hard drive, it's most likely in the process of failing, if it hasn't failed already. If you're in the process of troubleshooting a noise that you believe is coming from a mechanical hard drive, this needs to be confirmed. Once again, SSDs generally will not make any noise. User's may themselves be having problems because they're not fully aware of what's going on with the system, or they may simply be overloading the system. Some typical examples of user unawareness might be the following: • The hard drive is running low on free space. • The user is running so many applications simultaneously that the system can't handle them. • The system doesn't have enough memory available to handle the operating system and applications. • The user has applications running in the background they aren't aware of. • An external drive, such as a backup drive, frequently needs to "wake up" slowing system access. These, and other user problems are highlighted in our section titled "User Problems" which you can access by clicking on the icon accompanying the name at the top of this page. Software itself can cause problems by using high levels of CPU power, using large amounts of memory, or making excessive use of the hard drive. Some of these are addressed in the "User Problems" section described immediately above. Aside from actual user installed software, the following system applications may be responsible for problems: • The mds (meta data server) process used primarily by Spotlight and Time Machine. • The Icon Services Agent, which generates fairly large amounts of icon data. • Safari Web Services and other browser plugins used to handle video content. • Time Machine backups • Automatic software updates and installations. • Anti-virus software The mds process was problematic primarily in Mountain Lion (MacOS 10.8) and earlier versions of the operating system. It can still be problematic on newer OS releases, but it seems to occur mostly in bursts. The mds process seems to have problems indexing certain types of files, and if a lot of them exist on a system, it might cause the system to bottleneck as it can be extremely drive and CPU intensive. The Icon Services Agent is a process that generates icons for every single user account on the system, even if they're redundant. When processing icon information, it can literally eat up gigabytes of data in a matter of minutes, and continue doing so until it's complete. When highly active, this process is CPU, memory, and drive intensive. The more accounts there are on a system, the more files it will generate. Icon Services Agent appeared in Mavericks (MacOS 10.9). Safari Web Services and browser plugins may be extremely CPU and memory intensive. In some cases it may not really be the fault of the browser, but an abusively written set of web pages that can have the effect of "summoning up" several instances of these processes simultaneously. In most cases, the problems will stop when the browser is completely terminated. If a Time Machine backup is occurring unknown to the user, it too can be responsible for a performance drop and delays. Versions of MacOS prior to Mavericks (10.9) had an icon that could be installed in the menu bar, and when a backup was occurring the icon would spin until the backup was complete, which would allow a user to know a backup was occurring. With Mavericks and later, the icon can still be placed in the menu bar but it no longer spins during a backup, and if the backup drive is out of sight or has no indicators to identify backup activity, a backup may be occurring without a users knowledge. If the system has been configured to perform automatic downloads of software updates, this too can cause some problems. Typically during the download, network activity will go up as will disk activity, and when the installation starts, both drive and CPU activity can skyrocket. Fortunately automatic updates can usually be controlled by the user. Anti virus software can be notorious for causing not only delays, but sometimes lockups. Some anti-virus software needs to be "told" which applications on a system are allowed to gain access to resources. Most come pre-configured with many application and the operating system itself already allowed to run, but when a new, "foreign" application comes along that the anti-virus software doesn't recognize, it can delay it, subject it to extensive monitoring, or even block it completely until the user "tells" the anti-virus software that the application is allowed to run on the system. Failure to read or write data to a drive can exist with respect to an entire drive, be limited to specific files on a drive, or be erratic in nature. We've never encountered a condition where a user could write to a file system but not read from it unless there's a bad cable or other hardware fault or the permission settings of files are completely obscure and abnormal. If the problem occurs with respect to an entire drive, the following common possibilities are: • Index corruption • An SSD is entering its final life stages • There is no space left on the drive • Bad blocks on an SSD or bad sectors on a hard drive • Bad permissions Every drive, whether an SSD or a mechanical hard drive, uses an index to find the location of files on the storage media. The index, which is usually a reserved area at the start of the drive, stores the locations of the blocks and sectors that make up all the files stored on a drive. When a user summons up a file for use, such as opening a document for editing, the operating system will interrogate the index to find the location of all the blocks on the storage media, fetch them, and then load them into memory for use. If index corruption of the file system has occurred to such a severe degree that the OS cannot make sense of the file system the effected drives or volumes may be rendered read-only, and in extreme cases, may not be readable at all. This can often, but not always be corrected using the "repair" mode of Disk Utility. Causes of index corruption can be attributed to bad sectors or blocks, faulty cables, the user isn't allowing the system to shut down properly, or a sudden power outage or surge. This type of problem is more prevalent in earlier versions of MacOS X (10.3 and earlier.) If the drive is an SSD, it's quite possible that the drive is entering its final life stages. In this case the drive may end up being marked "read-only" because it can no longer replace write cycle depleted cells with spares. How a specific SSD handles blocks during their final stages of life will vary considerably from manufacturer to manufacturer. If there is no space left on the drive, whether it's an SSD or a mechanical drive, it will become difficult, and in some cases impossible to write data to the drive. If no free space exists, you simply cannot write to that drive any longer, but such a drive should be readable, often with considerable delay. The solution is obvious: delete data you no longer need. If the problem appears to be related to files on the drive, it's typically caused by bad blocks on an SSD, bad sectors on a mechanical hard drive, or a faulty cable or other hardware problem. Problems of this nature will usually occur during a read operation, and if either type of drive is excessively damaged, it can also inhibit the ability of write operations. Write failures on a mechanical drive will usually indicate severe damage and most likely a drive in the process of failing, with write failures on an SSD usually indicating end of life for the unit. The exception to the previous statement occurs when the problem is caused by a bad cable or other hardware problem. If the problem is erratic in nature, it's most likely a faulty cable or connection in the system. Another problem with failed read or write operations on files or folders may be related to the permissions assigned to a file or folder on the system. If this is the case, it may not be a drive or system problem, but rather the way the operating system works. Mac OS X is a Unix based system, and Unix based systems use varying permission levels for the type of user using the system. Not all users have access to all files on a system. To understand this limitation, it's best to review the various resources available regarding OS X for further details. It's also possible that an application may set permissions in such a way that the typical user is prevented from accessing them. If the file system frequently needs to be repaired by Disk Utility, it's most often caused by a user failing to shut the system down properly or ejecting drives incorrectly. However, if the system and drives are being treated properly, it usually implies that files in the system are "disappearing" because of the presence of developing bad sectors in a hard drive, bad blocks on an SSD, or there's a cable problem. These are described in the preceding sections regarding problems with bad/weak sectors on a mechanical hard drive, bad blocks on an SSD, and periodic delays caused (typically) by failing cables or connectors. When running applications lock up for no apparent reason, the chances of it being drive related, software related, operating system related, hardware related, or user related, are generally about equal. There is no such thing as a "perfect" operating system - all operating systems have bugs, as do all applications. Users may likewise be unaware of the fact that they're taxing the system CPU, memory, or even available drive space. A drive or its supporting components should really only be considered a suspect after the others have been eliminated as potential problem makers. Problems with the logic board or some of the components it supports, such as the Airport card may also cause these problems. See the sections above regarding problems with hard drives, SSDs, and cables. If a drive is no longer recognized by the system or a boot drive is recognized by the system but it will no longer boot, it will likely be one of the following: • A drive failure • Damaged boot regions • A failure in the system components supporting the drive, • A system configuration setting that's corrupt • An operating system bug • An incompatible interface or outdated drivers (if installed, and generally external units only) • The OS can't understand the drive format It's rarely possible for this type of problem to be a user induced problem, unless they've ejected the drive improperly, or in the cases of external drives, forgot to turn them on and/or plug them into the system (it happens!!) A hard drive, whether SSD or mechanical hard drive, can suffer from a controller card failure, which will make the drive invisible to the system. Such a drive will not show up on any system, regardless of its platform. For mechanical drives, if this has occurred, it's possible the drive may still spin up and appear to turn on, but fail to communicate with the system. In some cases the supply section of the controller card may fail and not provide adequate power to the drive itself, making the drive appear like it's not even connected to the system. Most drives use the very first sections of the media for boot code, index files, and other critical files. If bad sectors or blocks develop in these areas, the drive almost certainly won't be seen by the system. If the command line version of Disk Utility is used to list the drives, such a drive may show up as a single entity, such as "disk0" with no accompanying volumes or support partitions. Severe damage to the index files may have a similar effect. If there's a problem with the system components supporting the drive, it's usually, assuming the problem isn't cable related, the logic board. This will need to be properly diagnosed by a qualified individual. It the drive is external, it may be a lack of power to the drive (if it's port powered), a failure of the drive's supply, or a failed interface card in the external hard drive's enclosure. Note that if the interface card on an external hard drive fails, the data on the drive inside the enclosure may still be in good working order. If there are operating system bugs, configuration setting problems, and/or driver problems, a user may experience one or more of the following: 1. The drive isn't seen by the system. 2. The system appears to lock up when waking from sleep. 3. On reboot, the user is greeted with a white screen and no apparent system activity. If the system uses manufacturer provided drivers, make sure they are up to date and ensure that what you're witnessing is not a common problem with the drive. A number of these problems might be corrected by opening up System Preferences, selecting the Energy Saver option, and then make sure that the "Put hard disks to sleep when possible" option is unchecked. You may wish to look at the web pages in this set of articles regarding internal and external hard drives (see the links at the top of the page to access them) because some of the resets may correct the problems. If you're convinced the problem is related to a bug in the operating system, you should contact both the manufacturer of the drive and Apple to see if a workaround exists. When the operating system boots, the first thing it tries to do is look for hard drives so it can load the operating system kernel. Information about the drive to use at startup is stored in the volatile memory on the logic board. If this drive information is corrupt, you may be greeted with a white screen, and it may eventually show a prohibited (circle with a diagonal line through it) sign, a flashing question mark, or simply stay in a locked up position with the white screen. If this occurs, you will likely want to try some of the resets provided in the sections about internal and external hard drives. There are, however, cases where external devices, particularly USB devices, may misinterpret some commands from the operating system and leave the drive interface in an abnormal state. A way to work around this in some cases is to shut the system down, power down all devices, restart the system, and in a one-by-one fashion, start powering on the peripherals. This may or may not work, but it's easily done and doesn't require resets of any type. If the drive is visible to the system but won't boot, it usually indicates corruption of the drive's system files or a hardware failure. If it's related to the drive, it is often caused by the development of bad sectors in the drive, particularly in the first gigabyte of drive space. However, this can also be caused by problems with some of the system support circuitry for the hard drive, such as cables and connectors, or it may be a corrupt system setting. Improper shutdown by the user may cause these problems, as can problems with the logic board. If the problem is caused by a logic board failure, such as a blown video chip, the system may simply crash during a boot operation because the operating system kernel can't properly evaluate logic board components. Not all versions and formats of OS X are compatible with one another. Although the latest versions of OS X can typically deal with volumes from earlier releases, the opposite is not necessarily true. Core Storage, which was introduced in MacOS 10.7 (Lion) is not understood by versions of OS X earlier than this, and Lion based systems will not understand what a Fusion Drive is (Fusion Drives became available in Mountain Lion, or MacOS 10.8). Because of file system format changes, a user booted on a Mountain Lion volume may attempt to read data from a MacOS 10.10 (Yosemite) volume only to find it reporting I/O errors because the resource forks in Yosemite aren't understood by OS X versions earlier than Mavericks (MacOS 10.9). These aren't the types of I/O errors generated from a bad hard drive or faulty I/O cable, but they may be shown as such to any applications trying to access them. If a user attempts to boot from a drive that's formatted for a later release of OS X than the system can support, they may be greeted with a prohibited sign indicating the OS on that volume is not compatible with the system they are using. If a drive ejects itself for no apparent reason, it may be a bug with the operating system, but can also, particularly on external drives, be caused by a bad cable that's making the drive disconnect temporarily. On external drives, if the drive is port powered or the supply to the drive isn't adequate it may cause the drive to "drop from sight" as far as the operating system is concerned. If this problem is occurring on an internal drive, it's likely being caused by problems in the supply cables or connectors supplying power to the internal drive, but it can also indicate a logic board or system power supply problem. If a drive shuts itself off for no apparent reason, it can be an operating system bug, a problem with the hard drive's controller card or on board supply, or the supply to the drive. Although this problem will also indicate that the drive was ejected improperly, the difference will be that the drive will literally appear to be turned off. In some cases this may be an artifact of drivers associated with external hard drives. In most cases this usually indicates a hardware problem, meaning either the drive itself is failing or the supply to it is failing. This page is part of a set of pages which are identified at the top with a set of icons. If this article hasn't answered your questions, you may find answers in some of the other sections.