Main menu:

Site search

Categories

June 2017
M T W T F S S
« Apr    
 1234
567891011
12131415161718
19202122232425
2627282930  

Tags

Blogroll

A case of missing tombstones – data recovery from a Seagate Backup Plus Drive

external seagate hard disk recovery dublin irelandIn the world of data recovery, you come across many interesting cases. One such case was last week when a professional genealogist visited us with a failed external Seagate Backup Plus drive. Her job involves researching family lineage and history as far back as possible. It’s a laborious and time consuming occupation. Interviews have to undertaken, church records, birth and death certificates and newspaper archives all have to be methodically researched.

For this genealogist, part of her job involves travelling the length and breath of Ireland photographing tombstones in the church grounds and graveyards. She would number the photos (.JPEG files) and then input data, such as family name, graveyard name, townland and other relevant information into a Microsoft Access database. This was all stored on her Seagate Backup Plus USB external hard drive.
When she connected the drive to her laptop and heard a strange noise emanating from inside the disk she knew something was wrong. Her son-in-law, an IT manager kindly offered to see what he could do. As a seasoned pro, he instantly recognised that the clicking noise indicated a serious mechanical problem with the disk. Having used our hard disk data recovery service before, for his own organisation, he recommended that she should contact Drive Rescue.

We removed the disk, a Seagate Momentus 7200.5 500GB, from its plastic shell and commenced our diagnostics. Heads 03 and 04 failed our disk-head read test. When these heads were attempting to read the Service Area on the platters they could not access any of the drive’s initialisation microcode thus causing the clicking noise. With multiple disk head failure, the best course of action to maximise the chances of a complete recovery is usually to perform a Head Disk Assembly replacement.

 
We had an identical Head Disk Assembly taken from another Seagate Momentus 7200.5 500GB in our storeroom which had the same (HDA) part number. This would be our donor drive. The failed drive was opened in our Class-100 clean room. We used a tool called a “head comb” (no, not the type that Boots sell…) which is a device specifically designed to safely remove a HDA from a hard disk. There are various types of disk head-comb customised for each hard drive brand. In this particular case, the head comb was designed for working inside Seagate Momentus 2.5” disks. It connects through the small hole on the head arm and can be secured into place using a locking pin. The underside screw which holds the heads in place was then removed. After the brake of the drive was disengaged, the HDA can be safely removed from the disk chassis by using an anti-static tweezers.

The donor HDA was inserted and all components reconnected. It was finally time to close the drive lid and initiate power to the drive. The drive spun into life, but this time no heads were being detected at all. Sometimes, this is normal disk behaviour if the HDA-securing screw on the underside of the drive needs torque adjustment. By tightening or loosening the screw with the turn of a Torx key, the torque pressure can be easily adjusted. In this case, we loosened the HDA screw by turning it around 180 degrees. We applied power to the drive again, and this time all the heads were detected. We then connected the drive to another recovery system to search for a volume on the drive. An NTFS volume showed up with a substantial number of. JPEGs and.MDB (Microsoft Access) files. These were all extracted onto a new drive. The user could now login remotely to our secure systems to view their recovered data.

To say that the user was happy would have been an under-statement. For them to retrace their steps in re-photographing tombstones and re-entering details into a database again would have been a costly, time-consuming and soul-destroying task.

Advanced data recovery of inaccessible or corrupt Outlook .PST and .OLM files

 

recover outlook email data dublin

 

Outlook is a popular email application for Windows and Apple platforms. For most users, it is more than just an email client as it performs as a calendar, contact manager and task organiser. It uses a message store file called .PST (personal store folder) as it’s native file format. We have been helping a number of our clients recently recover data from corrupt .PST files and recover .PST files from mechanically damaged hard disks. For some of these clients, their priority was the recovery of their emails. For others, it was their email, contacts and calendar.

Common Microsoft Outlook error messages

 
When Outlook opens with an error message about an inaccesible .PST it can be quite scary for the user and sometimes for the IT admin (who has no backup…). Common Outlook error messages in Windows include:

 

Cannot open your default e-mail folders. The information store could not be opened.
Cannot start Micorsoft Outlook. Cannot open the Outlook window. The set of folders cannot be opened. The operation failed.
The file C:\Users\USERNAME\AppData\Local\Microsoft\Outlook\Outlook.pst cannot be found.
Cannot start Microsoft Outlook. Cannot open the Outlook window. The set of folders cannot be opened. File access is denied. You do not have permission to access the file C:\Users\USERNAME\Documents\Outlook Files\Username.pst

Some of these error messages are indicative that your .PST file has gone corrupt.

To understand how to recover .PST files, it can be helpful to know about the file’s architecture and format. In all versions of Outlook pre-2003, all .PST files were in ANSI (American National Standards Institute) format. From 2003 onwards, all .PST files are now in Unicode format which allows for 64-bit block IDs and absolute file off-sets. Typically, a .PST file is constituted of three layers. These include the Messaging Layer, the LTP layer (lists, tables and properties) and the NDB (Node Database) layer.

The structure of .PST files

data recovery outlook pst file dublin ireland

The messaging layer contains the message store, folder structure, messages and attachments. The NDB layer contains the lowest-level information about the .PST file such as header file and metadata pages. The LTP ( Lists, Tables and Properties) layer collects file properties. Because of the .PST’s proprietory status, Microsoft have not disclosed the algorithmic functions which are used to maintain integrity between the layers (the lousers…).

Data recovery or repair of .PST files

There are many “PST recovery tools” available to buy on the internet. Most of these promise the beleaguered user in search of a quick fix a “one click solution”. Sometimes they do work. But most of these utilities cannot handle extensive LTP, NDB or messaging layer corruption. The SCANPST utility can sometimes be successful if there just low-level corruption. However, if the damage if more extensive Drive Rescue can help with a complete data recovery of your Outlook .PST file.

The .OLM connection (Outlook for Mac)

The relationship between Outlook for Windows and Outlook for Mac is rather tenous. In fact, in Microsoftland the developement teams for both aplications are not in the same building or even part of the same business unit. As a result, while on the surface they do share similarities, their underlying architecture of both applications is quite different. The common Outlook for Mac 2011 application has a measly upper size limit of only 2GB. When your .OLM file goes corrupt you can try rebuilding the file by using the Microsoft Database Utility or by rebuilding your identity. If all this fails, professional data recovery for your .OLM file might be needed.

How to mitigate against data loss with Outlook

Use IMAP instead of POP3 – If you’re using Outlook 2007, 2010, 2013 or 2016 try to use IMAP instead of the POP3 protocol for your email configuration.(POP3 was not really designed for era of mobile email use). Because IMAP syncs your .PST with your mail server, should something go wrong with it, – you should be able to access a copy of your emails on your mail server.

Don’t let your .PST file get too large. Outlook 2003 and 2007 has a theoretical file size limit of 20GB. From our experience, however, for these versions of Outlook, once you let a .PST go above 16 to 17 GB in size, Outlook can start to go awry and you run the risk of your.PST file imploding. For Outlook 2010 and 2013 the .PST file size limit is a very generous 50GB.

Use Microsoft’s PST Back-up Tool – This handy utility from Microsoft allows you to backup your .PST by using an Outlook plug-in. In turn, you can use its output .PST file for on-line or local backups. This can be extremely useful because when you have Outlook open, a substantial number of backup applications still refuse to back-up a .PST because it is a “file already in use”. Having a second up-to-date .PST file for back-up purposes is a great workaround to this common problem.

Use Outlook Office 365 – With this option your .PST is stored in the Cloud. This can mitigate the risk of data loss due to disk failure but does not protect against events like file corruption. For this, it is recommended that you use an Office 365 solution such as Cloudally or Backupify.

In the event that you or one of your users has lost access to their email store file, Drive Rescue offers an advanced .PST and .OLM data recovery service for corrupt files or files that are inaccessible due to disk electronic or mechanical faults. Call us on 1890 571 571.

Data recovery from 128gb Lite-On IT mSata SSD from Dell Laptop

data recovery liteon ssd drive dublin irelandDrive Rescue is peforming data recovery from an increasingly large number of SSD drives. One driver for the popularity of SSD drives is the reducing cost of NAND memory. Another being the proliferation of light mobile devices – users now have an expectation that endpoint computing devices should be light and small. The success of NAND memory is self-evident. It has already killed off the floppy disk, 35mm photographic film and the one-inch hard drive. If you buy an ultra-portable laptop today, more likely than not it will have an SSD drive as its primary storage device. All MacBook Air laptops now have an SSD drive as standard.
SSD drives are not without their problems however. Take for example this Lite-On IT (pictured above) drive from a Dell laptop. The user of the said system discovered that it would no longer boot to the operating system. Using an mSATA adaptor, he tried reading the drive on another system but the drive was not showing up.
They delivered the drive to us – a Lite-On IT LMT-128L9M. The drive did register on our systems, but it’s ID appeared as scrambled hieroglyphics. Our diagnosis revealed that the drive had a corrupted microcode. So after some extensive searching we found an exact match microcode for this SSD which we then uploaded to our data recovery system. The drive now ID’ed but there was sill no data showing. Before we started any logical recovery, we imaged the drive to another one of the same capacity. Then using a hex editor we looked at every record in the file system for valid NTFS entries. We eventually found one promising looking NTFS partition which we rebuilt and then rebooted the drive . This time data appeared. We invited the client to login remotely to our secure systems to view his files. Word, Excel, Outook (.PST), Onenote (.ONE), .PDF and .RAF (Fuji Raw Image) were all successfully recovered for a very relieved and satisfied client.

Repair of burnt smooth motor-controller chip on Western Digital drive

fix western digital burnt smooth chip dublin ireland

Problem
We were recently sent a 1TB Western Digital (S-ATA) drive from Waterford. The user could no longer access their data and when they connected the drive to their Apple Mac they got a faint smell of burning.

Diagnosis

The user’s findings were correct. A burning smell could be detected withing 30 seconds of power-up. We proceeded to remove the PCB and it was immediately apparent that the problem was a burnt smooth motor controller chip. We used then used a multi-meter to test other components of the PCB. Luckily, these appeared to by okay.

What function does a motor controller chip on a hard drive perform?

A smooth chip or motor controller chip controls rotational speed of the platters, but also acts as a DC to DC converter.

Why do motor controller (smooth) chips fail?
Motor controller chips tend to be subjected to the more unregulated voltages from the PSU than other chips on a hard drive’s PCB. A short-circuit of any kind can cause the transistors to heat-up and subsequently burn out.

What is the fix?
Burnt motor-controller chips are a very common occurrence which we see on Western Digital and Seagate drives. The fix usually involves replacing the motor controller chip itself or replacing the whole PCB. If you go for the latter option, you will also have to tranfer over the drive’s unique adaptives information stored on the drive’s BIOS chip or NVRAM chip (Hitachi).

What is the success rate like?
Success rates for this type of data recovery is excellent. However in a rare number of cases, if the drive’s pre-amplifier has been damaged by the short-circuit, this might need replacing also. The 1TB Western Digital drive whose PCB is in the photo above was recovered completely and the customer had all his data back in 3 days!

Recovery of Solidworks files from Dell RAID 5 Server

 

solidworks data recovery from dell raid 5 server dublin cork limerick

We recently helped a Dublin engineering company recover data from their Dell PowerEdge T310 RAID 5 server. Last week they reported to their IT support admin that their server would no longer boot-up. He performed diagnostics on the drive and discovered that one of the disks (a 1TB Seagate Barracuda) had failed. Thinking it was not that serious, he removed the failed disk and replaced it with a 1TB hot-spare. Under normal circumstances, the RAID controller (in this case a Dell Perc) should make a block-for-block copy to the hot spare which would then become an active drive after the RAID rebuild process. However, after he initiated the RAID rebuild process and after a 6 hour wait he was getting nowhere. The RAID array would not rebuild. Knowing the limits of his expertise and having used Drive Rescue successfully before – he called us for assistance.

We labeled all the drives and noted the disk order. We then removed the drives from the server and then imaged each drive. Working with the drive images we used our specialised tools and experience we determined the block size, parity strip size and RAID offset.

After several hours work, we now had a volume which we could extract data from. In this particular case, their most important data to the client were Solidworks files, namely assembly files  (.sldasm), drawing files (. slddrw) and image files (. sldprt). These were all extracted and checked for integrity. The client was then able to remotely login to our secure systems to view these files for themselves. Having got confirmation that all of their files were there, we proceeded to extract them to a 2TB USB portable drive which we delivered to a more-than-happy customer.

Data recovery from Dell PowerEdge R805 rack server with integrated RAID controller

raid data recovery dublin cork galway ireland  limerick  ireland

 

It was the afternoon before a major match when a well known Dublin stadium discovered that their Dell PowerEdge R805 server running Windows Server 2003 was no longer booting. Upon boot-up they were greeted with a “disk error occurred” error message. They tried to perform a RAID rebuild but the rebuild process would not even initialise for them. Their server was running MySQL which was tied to their retail POS software. Without this server, most of their tills for their restaurants, shops and bars would be non-operational. Without some of their receipts to customers would have to be hand-written. They needed a solution and needed one fast.

With such a limited time frame, Drive Rescue pulled out all the stops to help them. We went onsite to perform the initial diagnosis. The drives were configured in RAID 0 (known colloquially as “scary RAID” for it’s total lack of redundancy when it does fail).

dell hp fujitsu raid data recovery ireland

We removed the 2 x Dell Constellation 500GB entreprise-class S-ATA drives and brought them back to our lab. Drive 0 was in rude health. Drive 1 had over 29,000 bad blocks which had to be reallocated. Using a hex-editor, we manually rebuilt the MBR. Then we went performed the RAID rebuild process including finding parameters such as block order, block size, stripe size and RAID offset.
With the volume now rebuilt and extracted onto two new 500GB drives. We then configured it so that it would be bootable again. Understandably, with a major match looming in the stadium later that day the client would not have enough time to re-install and the operating system and all ancillary software.

Using a hex-editor, we manually rebuilt the MBR. Then we went to perform the RAID rebuild process, including finding parameters such as block order, block size, stripe size and RAID offset.

With the volume now rebuilt and extracted onto two new 500GB drives. We then configured it so that it would be bootable again. Understandably, with a major match looming in the stadium later that day the client would not have enough time to re-install and the operating system and all ancillary software.

dell poweredge server rebuild dublin ireland

After burning some midnight oil, the following morning we delivered the two disks containing the recovered volume and connected them to their server. The power button was pressed and some reassuring beeps were heard and finally the Window Server 2003 logo appeared followed shortly by the desktop screen. Sigh of relief all round! Now their retail operations would not be interrupted and they could safely scupper their plan B of writing out receipts with paper and pen!

Data recovery from LaCie Big Disk 1TB RAID 0

lacie big disk data recovery

We recently helped a management consultant from Co. Cork recover data from his failed Lacie Big Disk configured in RAID 0. He bought the drive 4 years ago for personal and business use, and it worked great for him. However, last week when he went to turn it on, it was not recognised by his iMac or MacBook Pro computer. He look it to his local computer repair shop. They removed the disks from their enclosure, and they discovered that Drive “0” was not operational. They referred him to Drive Rescue data recovery

lacie raid disk recovery ireland

We opened the case and found 2 x Samsung HD501LJ 500GB disks inside. Disk 0 was clicking when initialised whilst Disk 1 seemed to be healthy. The problem with a RAID 0 configuration is that the data is spread across two disks. Thus, having one disk operational does not mean that you still have half your data because it is spread across the two disks evenly. For example, if you have a .JPEG file, 50 percent of its constituent bits will be stored on Disk 0 and the other half on Disk 1. We examined Disk 0 further. Heads “2” and “3” had failed. This would require a head-disk assembly swap. Because Disk 1 was identical, it means we could transplant the HDA from Disk 1 to Disk 0. But first, we would have to image Disk 1.

 

 
Once the imaging for Disk 1 completed and we had verified it’s integrity, we brought the two disks into our cleanroom from their “operation”. Firstly, this involved removing the “good” HDA from Disk 1. This is an intricate procedure which requires skill, experience and a steady hand. Once this HDA was removed from Disk 1, it was now time to remove the defective HDA from Disk 0. Now both HDA’s had been successfully removed.

 
Now it was time to start the “transplant”. There are a number of criteria here which must be met for a successful head disk assembly swap. The transplanted HDA must be aligned perfectly with the disk platters. If wrongly aligned the “flying height” of the heads will be too low or too high. In both cases, this can result in the non-reading of some or all of the disk tracks. Another important factor to mention is to torque the restraining screws correctly. If there is not enough pressure, the platters can warp. Too little pressure and the syncing between heads and platters will be off-kilter.
With the replacement HDA from Drive 1 now successfully transplanted to Drive 0, it was not time to image it. This took approximately 3.5 hours.

 
The second stage of this data recovery would involve rebuilding the RAID array of the two disks so that original volume could be recreated. This involved finding the block size, the RAID offset and the block order. These parameters were found using a hex editor.

 
After some hours of calculations the RAID parameters were found and the original volume name now reappeared. Word, Excel, PDFs, .JPEGs, and Sage Accounts files were all extracted onto a USB external drive and dispatched to our more-than-happy client in Co. Cork.

Giving Back @ Drive Rescue

drive rescue data recovery dublin ireland giving back

Homelessness in Ireland is a problem which has been in the media spotlight recently. But homelessness has always been a problem. We realise that the causes are deep-rooted and multi-faceted but some help is better than no help.

The charity we decided to donate to this year was the Capuchin Day Centre in Dublin 7. It serves breakfast to around 250 people each day and dinner to around 450 and 520 people. The operating costs of the centre are around €2.3 million but the State only provides €450,000. The rest comes from fundraising. It was enlightening to see the quality service offered by the centre and to sit down with Brother Kevin (head of the centre) to discuss the homelessness problem. His centre does an admiral job in providing food, warmth, showers, clothing, companionship and heathcare to those less fortunate. So it was a real honour for Drive Rescue to be able to make a contribution to this very worthwhile cause.

The Drive Rescue team would like to wish everyone an enjoyable Christmas and a happy 2015.

Data recovery from an Intel 1.8” SSD drive

data recovery from SSD dublin ireland

Problem
Recently, a user from Athlone was using his HP Elitebook laptop connected to his mains power. (His battery was no longer able to hold a charge). Unfortunately, when rushing to his answer his phone, he accidently tripped over his laptop’s power lead and almost instaneously his system powered off. No worries, he thought. This happened to him before and did not pose much of a concern for him. After a sudden shutdown before, his laptop booted up successfully.
He finished his phone call and went to switch his laptop back on. But this time he got the “operating system not found” error message. He turned it off and back on again. But still, the same error message re-appeared. He brought the laptop to his local computer repair shop. They removed the Intel SSD hard drive from his system. It appeared on their system alright, but it was only detected as having a capacity of just 8Mb! They tried running some data recovery programs on the drive, but any data was proving to be elusive. They referred him to Drive Rescue data recovery.

Using our eqiupment, specialised to recover data from SSD devices, we discovered that the logical-to-phyical table of the drive had gone corrupt. The role of this table is to translate LBAs (logical block addresses) to physical addresses (chip number and page number of chip). When there is the sudden loss of power, this layer often goes corrupt.

Solution

We first put the drive into safe-mode and then used our equipment to find the corrupt L2P module. After we found the corrupt module (containing the corrupt table), we then had to upload a “good” module to match the model number (SA1M160G2HP) and firmware version (02HA). After some extensive searching of our firmware database, an exact-match module was found and uploaded onto the RAM of our recovery system. After installation, a new volume appeared, but it was now only showing 48GB. This was certainly not the size of the original volume. Something else was awry. We decided to change the recovery mode of our system from “tech-command” to “translator-table” mode. With this parameter changed, the whole 160Gb volume appeared. The downside of TT mode is that it automatically changes the read mode from UDMA to PIO which gives a data transfer speed of only 4Mbps. But slow-and-thorough is always better than fast-but-incomplete.
All of the client’s data was recovered, namely his .PST file, his .ACCDATA file (Sage) and his Microsoft Office data files onto a brand new USB external hard drive. The client has informed us that he is now going to decommission his old HP Elitebook which gave him such a nasty surprise in favour of a new MacBook. An early Christmas present to himself.

Recovering data from a Seagate 7200.11 disk with bad sectors

recover data from disk with bad sectors dublin ireland recovery service2

One of the most insidious types of disk failure on conventional hard drives is due to bad sectors. A bad sector is a sector on a hard disk that cannot be read, written or corrected by the drive’s ECC (Error Correction Code) mechanism. Typically bad sectors develop slowly over time and the average computer user often gets no indication that there is a problem with their disk.

Bad Sector Symptoms

A disk with bad sectors will sometimes show these error messages.

You need to format the disk in drive F before you can use it ” (Windows operating system)

The disk in drive E is not formatted. Do you want to format it now?” (Windows operating system)

The disk you inserted was not readable by this computer” (Mac operating system)

recover from bad sectors on apple mac disk drive

Modern drives more likely to develop bad sectors

Modern multi-terabyte drives using perpendicular recording are more likely to develop bad sectors compared to their smaller capacity brethren. This is because they use higher areal densities (more bits are squeezed into the same space), the track widths are narrower and their disk heads fly lower. These attributes make the drive’s signal-to-noise ratio decrease. As the SNR ratio decreases, the likelihood of bit-errors developing increases. However, this is somewhat compensated for by manufacturer’s use of more sophisticated ECC algorithms.

Other factors which increase the probability of bad sectors:

Age – As disks age, the probability of bit-flip and other magnetic distortions occurring increases. Hard disks are not the only magnetic storage medium to deteriorate with age. For example tape storage is notorious for age-related degeneration.

Thermal Asperities – Tiny particulates of contaminants inside the drive can cause a phenomenon known as thermal asperity. Typically, it occurs when a giant magneto-resistive drive head collides with a contaminant on the disk platter. As a result, heating will occur on the disk head and platter surface. This causes some sectors in the affected area to become unreadable.

Dirty Power – If the host system of your hard disk is delivering intermittent over and under voltages to your HDD, then the probability of bad sectors increases. Read/write heads need a stable power supply to perform read, write and erase functions properly.

Recovering data from a drive with bad sectors

In a small number of cases where the disk has relatively few bad sectors, commercially available data recovery programs can be successful in recovering your data. I emphasize a small number of cases because, as soon as these programs hit contiguous bad sectors, they will often get stuck or freeze the host computer. These programs are really only designed for disks with relatively few bad sectors. If the disk has any underlying read/write head issues, repeated retries can precipitate failure of one or all of the disk heads. Users who torture their drives by using these DIY programs on damaged drives risk permanently losing their data.

The right equipment and skills

If your data is in any way important, the best way to recover data from a disk which has extensive bad sectors is to take it to a professional data recovery company. They will have the equipment to deal with bad sectors. But, having the right equipment is not always enough. In the same way that a kitchen fitted out the with latest cookers, mixers and knives does not make a Michelin starred chef – the data recovery technician must have experience, skill and insight. Some bad sectors recoveries are straightforward procedures. Other cases will be more complex.

Data recovery from a Seagate Barracuda 7200.11

Take for example a client we were helping last week. A vetinary surgeon in Waterford was using a HP desktop PC with a Seagate Barracuda S-ATA 7200.11 as a server for his x-ray machine. Recently, the drive became inaccessible and his local IT support company were unable to retrieve his files. They were of vital importance to him. Without them, he would have to schedule appointments with dozens of his customers again in order to x-ray the animals for a second time. This would have imposed a huge time burden on his staff and would have been damaging to his reputation.

Our preliminary scans revealed that his disk had extensive bad sectors. To complicate matters, most of these bad sectors were in the inner tracks. This can be the worst place for a drive to develop bad sectors as the Master File Table file is usually stored here. This file is important because it acts as an index for the whole drive.

The data recovery from this drive involved a number of steps. Using our equipment, we disabled SMART on the drive. (In this case SMART only gave the user a warning after the drive failed to boot up). During the recovery process, SMART will actually hinder the recover process. Next, we put the drive into PIO mode instead of UDMA mode. Recovery from disks with bad sectors is made easier in this mode as it enables much better quality reads. Then we set the read time-out and read block size. Optimal read time-out will vary depending on the extent of the damage. Sometimes a time-out of 1200 millie-seconds is needed for reading from badly damaged sectors. In this particular case, our setting of 550 millieseconds proved optimal. We set the read block size to 60 sectors. The means our equipment would read 60 sectors at a time. This setting proved most suitable. Two hours into the process, the MFT file was successfully copied. The rest of the data took nearly 16 hours to copy.

xray dicom

Our process was successful. All of the x-ray files in .dicom format were recovered. We are able to use the excellent MicroDicom reader (created by Simeon Antonov Stoykov) to verify the integrity of the images. Our client was able to securely login to our systems to view his recovered files. A lot of labradors, poodles and cats had been saved a second trip to have their innards photographed!

Protection against bad sectors

For standalone disks, there are some inbuilt protections against bad sectors: Error Correction Code and SMART. ECC is designed to detect errors and try to remedy them. These are merely safeguards and nothing else. Hard disk manufacturers are sometimes a bit reluctant to over-burden their disks with ECC mechanisms for fear it creates too much of an operational overhead on the disk’s performance. Then there is the problem with SMART. Theoretically, this mechanism is designed to detect bad sectors and then report them to the user. But because a SMART alert entitles the user to an RMA (Return merchandise authorisation), manufacturers have set the SMART threshold extremely high. Often, the user only gets notified by a SMART alert when the drive is at an advanced stage of failure or has already failed.

Myths about bad sectors

A pervasive myth among general computer users is that only disks in Windows systems develop bad sectors. The reality is bad sectors are operating system agnostic. Whether a disk is running the latest version of Mac OS X, Linux or the latest version of Windows; it is still possible for the disk to develop bad sectors.

Hard Disk Sentinel to the rescue

There is an wealth of “disk health” programs available that promise to monitor the health of your disks. Some of these are adequate, others just do what SMART monitor does already. However, one disk health application which stands out for its accuracy is Hard Disk Sentinel. It provides some predictive indication if bad sectors are developing on your disk.

In the context of RAID arrays, bad sector management becomes much more sophisticated and would merit another another blog post to detail its intricacies.

Last word

Ulimately, the best prevention against bad sectors is backing up your data on a regular basis. Our vet client now uses an external hard drive to backup his x-rays. We recommended to him that he use a quality online backup service as an extra layer of protection.