Lenovo ST650 RAID 5, 8 SAS Drives, 2 Failed: Files Recovered After Another Lab Spent Two Months

This Lenovo ST650 RAID 5 two failed drives data recovery case study covers a VMware server with eight 900GB SAS drives. Another company had worked on it for two months without getting the data back. We imaged every drive, rebuilt the worst one with donor heads, worked out a RAID layout no software could detect, and the customer’s files came back after more than three weeks.

Sound familiar?

• Two drives in a RAID 5 server have failed and the volume will not mount.
• RAID recovery software cannot detect the array or its settings.
• Another company has had the drives for weeks with no result.
• The drives show SMART warnings or bad sectors.
• The virtual machines on a VMware server can no longer be reached.

Same problem? Call 1300 444 800 or request a free assessment.

What came in: a Lenovo server another company could not recover

The server was a Lenovo ST650 running VMware, with eight 900GB SAS drives in a RAID 5 array. Two drives had failed. RAID 5 survives one failed drive. It does not survive two.

The customer had already sent it somewhere else. That company quoted more than US$40,000 and worked on it for two months, but could not get the data back. Lenovo themselves then pointed Iurii to us, and the drives came to our main data recovery lab in Rhodes, NSW.

Payam Toloo, who founded the business in 1998, called it one of the most complex jobs the team has taken on.

What did the drives show when we tested them?

Every drive went on the Atola Insight Forensic first. All eight showed SMART warnings and bad sectors, not only the two that had dropped out of the array.

So all eight were cloned before anything else was tried. Six came back almost perfect. One took 36 hours. The worst took nearly five days.

Cloning first matters on a job like this. Each read of a failing drive can make it worse, so the work is done once, on the images, and the original drives are left alone.

Why could RAID recovery software not rebuild this array?

With the images ready, the real puzzle began. To rebuild a RAID 5 you need the disk order, the block size and the parity layout. This one was unusual.

UFS Explorer could not detect it. ReclaiMe ran for 24 hours and failed. No recovery software we tried could work out the settings, so they had to be worked out another way.

The Lenovo ST650 RAID 5 recovery process

Payam turned to Claude Cowork, an AI tool running locally on his own machine with strict privacy controls. Over days of back and forth, it helped establish two things no off the shelf tool can even attempt:

• One drive had failed long ago, so its data was stale and it had to be left out of the rebuild.
• Another drive needed patching with data from a second drive to repair it.

In parallel, the team rebuilt the worst drive by hand. Donor heads were fitted and the drive was re-imaged on a DeepSpar Disk Imager.

This was real trial and error. Then came the low point. Payam found that one of the parity settings he had been given was wrong, and days had been lost chasing it. In the video you can see the frustration: so close, and nothing working.

He did not give up. When Claude’s newest model, Fable, was released, he started fresh. It caught the mistake, and then pinned down the exact partition inside the VMware volume.

Lenovo ST650 RAID 5 data recovery results

After more than three weeks, the customer’s files appeared. The relief in the lab is on camera.

• 8 SAS drives of 900GB each imaged, every one with SMART warnings and bad sectors
• 6 drives cloned almost perfectly, one in 36 hours, the worst in nearly five days
• 1 drive rebuilt with donor heads and re-imaged
• 1 long failed drive left out, 1 drive patched from another
• Files recovered from the VMware volume after more than three weeks

Everything is being saved to a new drive, with a file list and a video preview ready before the customer collects and pays in person.

Why does a RAID 5 with two failed drives need a specialist?

When a second drive fails, RAID 5 has no spare parity left, and the controller can no longer present the volume. Rebuild attempts, reinitialising the array or running software on the original drives can overwrite what is left.

This job needed four things at once: imaging equipment that can read drives with bad sectors, donor heads fitted to a SAS drive, the ability to spot that one drive was stale, and the patience to work out a layout no software recognised. If your server has lost two drives, power it down, do not rebuild or reinitialise, and send the drives in, labelled in their slot order if you can.

For more on how we handle arrays and servers, see our RAID data recovery service page.

Why People Choose Payam Data Recovery

When a RAID, NAS or server fails, the array has to be rebuilt correctly the first time, without putting the original drives at further risk. That is the work we do every day.

• We are a specialist data recovery lab, not an IT support company. We have been recovering data since 1998, from over 300,000 devices.
• Over 2,000 IT companies and computer shops across Australia send us their customers’ complex cases, and data recovery labs and IT companies in more than 50 countries have sent us their difficult cases over the past ten years.
• Every drive is tested and cloned first on professional tools such as Atola Insight Forensic, the PC-3000 and the DeepSpar Disk Imager, and the RAID is rebuilt from the clones. A drive that needs a head swap is opened in our ISO 5 certified laminar flow cabinet, with over 30,000 donor parts in-house, including around 20,000 donor hard drives.
• Emergency service for businesses. Many of our business and IT customers choose it, because every day a server is offline usually costs more than the difference in price. 24/7 emergency service is available as a premium option on 0404 872 926.
• Your data stays private. Our data recovery engineers open only a few random, non-sensitive files to check the recovery worked, and we will sign your NDA on request.
• See your files before you pay. You tell us which files matter most. Before you pay for the recovery, we’ll send you a list of your recovered files so you can check them. If any of the files you listed are missing, tell us and we’ll search again.
• There is no card surcharge. Pay by Visa, Mastercard, American Express, Afterpay, Zip or bank deposit (EFT).
• Rated 4.9 from more than 2,300 Google reviews.

We are not the cheapest. We are the best in Australia at this work. We know what to do, how to do it, and we will do our very best for you. Your best chance of getting your data back is with us.

RAID and server data recovery pricing and services

Every RAID and NAS job starts with a free assessment and a written quote, so you know the exact price before any chargeable work. Complex multi-drive server cases are quoted individually, and current prices are on our RAID and NAS data recovery page. Priority and Emergency are faster and cost more.

If you need an answer fast, an urgent assessment and written quote within 2 to 3 hours is available for a fee on RAID and NAS jobs with two or more drives. Priority and Emergency service levels shorten the turnaround once the quote is approved.

How our process works for international customers

International customers are always welcome. The easiest way to reach us from overseas is by email at help@payam.com.au, which works in every time zone. We assess the drives, send a written quote, and once the data is recovered you receive a full file list and video preview before paying.

Frequently asked questions

Can a RAID 5 with two failed drives be recovered?
Often, yes. The failed drives are imaged first, with donor heads or other repairs where needed, and the array is rebuilt from the images rather than the originals. In this Lenovo ST650 case every drive had bad sectors and the files still came back. The outcome depends on how much of each failed drive can be read.

Another company already tried and failed. Is it too late?
Not necessarily. In this case another company worked on the server for two months without success, and we still recovered the files. What matters is what the job needs: imaging of failing drives, donor parts, and working out the RAID layout by hand when software cannot detect it.

What if RAID recovery software cannot detect my array?
Then the settings have to be worked out another way. Here UFS Explorer could not detect the array and ReclaiMe ran for 24 hours and failed. We established the disk order, block size and parity ourselves, left out a long failed drive and patched another from a second drive.

How long does complex RAID recovery take?
It depends on the drives. Most straightforward arrays take days. This one took more than three weeks, because the worst drive needed nearly five days of imaging plus a donor head rebuild, and the RAID layout had to be worked out by trial and error before the VMware volume would open.

Is the assessment free, and is it no fix, no fee?
Yes. The assessment and quote are free, and for most faults there is nothing to pay unless we recover your data. See your files before you pay. You tell us which files matter most. Before you pay for the recovery, we’ll send you a list of your recovered files so you can check them. If any of the files you listed are missing, tell us and we’ll search again. The one exception is cleanroom and other high risk work, where part of the quoted price is a non-refundable attempt fee paid upfront so we can begin. That covers the data recovery engineer time and donor parts we commit whatever the outcome. If an attempt fee applies to your job, it will be stated clearly in your quote before you approve anything.

Should I try to rebuild the array myself first?
No. Power the server down and leave the drives as they are. A rebuild or reinitialise on a degraded RAID 5 can write over the data that is still there. Label each drive with its slot number, then call 1300 444 800 or request a free assessment.

Can I send a server from overseas?
Yes. International customers are always welcome. The easiest way to reach us from overseas is by email at help@payam.com.au, which works in every time zone. You send the drives, we assess them and send a written quote, and you see a full file list before paying.

About Payam Data Recovery

Payam Data Recovery has been recovering data since 1998, with over 300,000 devices recovered. All work is done in Australia by our own data recovery engineers, never outsourced overseas. Complex cases go to our main data recovery lab in Rhodes, NSW, which has ISO 5 laminar flow cabinets, and we have full service labs in Sydney CBD, Melbourne and Brisbane. We will try our very best for you. There is never any guarantee with data recovery, but coming to us gives you the best possible chance.

Request a free RAID data recovery assessment

Related case studies

• Lenovo server RAID 5 rebuild failed: 10 VMs back in 24 hours
• VMware ESXi server, 8-drive Dell SAS RAID, recovered in 12 hours
• RAID 5 head swap on a Seagate IronWolf drive for an architecture firm

Video transcript

Hi, it’s Mike, and today I’m in Sydney, Australia, at Payam Data Recovery with owner Payam Toloo, for one of the most complex jobs his team has ever taken on. Eight 900 gigabyte SAS drives from a Lenovo ST650 VM server, a RAID 5 with two failed drives. Another company had quoted over $40,000 and worked on it for two months, but could not get the data back. Lenovo themselves pointed the customer to Payam.

On the Atola Insight Forensic imager, every single drive showed SMART warnings and bad sectors. Payam cloned all eight. Six came back almost perfect. One took 36 hours, and the worst took nearly five days. With the images finally ready, the real puzzle began. Working out the disk order, the block size, and the parity of a very unusual RAID. UFS Explorer could not detect it. ReclaiMe ran for 24 hours and failed. No recovery software on the market could work out the settings.

So Payam turned to Claude Cowork, an AI tool running locally on his machine with strict privacy controls. Over days of back and forth, it worked out that one drive had failed long ago and should be left out, and that another needed patching from a second drive to repair it. Something no off-the-shelf tool can even attempt. In parallel, Payam’s team rebuilt the worst drive by hand, fitting donor heads and re-imaging it on a DeepSpar Disk Imager.

This was real trial and error. Then came the low point. Payam realised one of the RAID parity settings the AI had given him was wrong, and that he had lost days chasing it. You can see the frustration on his face here. So close, but nothing working. He did not give up. When Claude’s newest model, Fable, was released, Payam started fresh. It caught the mistake, then pinned down the exact partition inside the VMware volume. After more than three weeks, the customer’s files finally appeared, and the relief and the celebration here says it all.

Payam is now saving everything to a new drive with a file listing and video preview ready before the customer collects and pays in person. The cost? Around 6,000 Australian dollars, where another company had quoted more than 40,000 US dollars and still could not do it. This is the kind of case Payam Data Recovery is known for, the complex RAID and server recoveries that other labs give up on.

Based in Sydney, Australia, and trusted on cases like this since 1998, Payam and his team welcome customers worldwide who have a difficult RAID, NAS, or server failure. If you have a complex case, this is who you send it to. They offer a free assessment and quote, they love a challenge, and they don’t give up. Thanks for watching.