Diagnosing MacBook Pro Hard Drive Cable Failure

[l008com]

I've been fixing Macs professionally (not through Apple) for well over a decade now. By far the most common thing I do is replace failing hard drives. While Apple's own tools (hardware test, Disk Utility) do almost nothing to positively confirm a failing hard drive, there are several other tools (SoftRaid, SMART Utility) that can read the SMART data and tell you for sure that a drive is failing. So unlike the Genius bar, I can quickly diagnose a bad hard drive without having to infer the problem through symptoms.


But one problem that I've noticed really being on the rise lately, is a failing hard drive cable in MacBook Pros. That's the little flat ribbon cable that connects the hard drive to the logic board. It also connects the sleep light. This is a much harder problem to diagnose, because it generally acts just like a system with a failing hard drive, but the drive will pass a real SMART test with flying colors. You might think "come on, how often does a simple cable really go bad?" Well, I've never had a traditional desktop style SATA cable go bad. But I've replaced a lot of these drive cables and I'm doing another one tomorrow. I never really know if it's going to fix the problem or not though...


... because, there's no way to directly confirm the cable is the problem. I know all about Apple service procedure. I know the official way to diagnose a problem is to start replacing parts until the computer is fixed, and that last part was your problem. But that's not how things go in the out-of-warranty world of independent Mac repair.


So I got to thinking. Computers with bad cables don't crash, or freeze. They just hang for a while randomly, then come back. JUST like a hard drive with lots of bad blocks. So what could be causing this? The only thing I can think of is SATA bus errors that only happen intermittently (on a scale relative to the amount of i/o calls a hard drive does). If that is the case, then there must be SOME kind of utility somewhere that can see these errors? There are plenty of utilities out that that have fancy 3D interfaces, but do absolutely nothing (tech tool etc). But there are a few tools that the Pros use that actually DO stuff. DiskWarrior and SMART Utility are the main tools. If I could find something that can somehow easily and definitively sniff out these bad SATA cables, that would make my life a whole lot easier.

 

[Archer1212]

When I worked at the genius bar, If the system booted to a flashing folder icon that was 99% it was the HDD flex cable. Basically any issues that seem HDD related for any 13" macbook pro, we eliminated the cable and continued testing. We would eliminate the cable by either using a firewire, Thunderbolt, or USB hard drive dock, or we would plug in a known good cable.

 

[The_Moves]

Are you the first person to see these Macbooks? Could the previous tech who worked on them have damaged the cable?

And yes, Computer with bad cables do crash and they do freeze - at least computers running a UNIX/Linux operating system. UNIX/Linux are far less tolleratnt to losing their disk than windows. I know from a VMWare/ESXi perspective, when we would have a NFS datastore outage, the Linux VMs would kernel panic where the Windows VMs would start working again once the datastore appeared and would act as if nothing happened. We had to increase the timeout policy for the kernel panic to allow the Linux VMs to survive outages for a specific period. If that time period threshold was breached, the VM would panic.

 

[Dan]

I swap out flex cables atleast once or twice a month...

 

[l008com]

When I worked at the genius bar, If the system booted to a flashing folder icon that was 99% it was the HDD flex cable. Basically any issues that seem HDD related for any 13" macbook pro, we eliminated the cable and continued testing. We would eliminate the cable by either using a firewire, Thunderbolt, or USB hard drive dock, or we would plug in a known good cable.

But if you're eliminating the flex cable by booting over firewire or ethernet, you are also eliminating the hard drive. Which leads you right back to the problem of: is it really a bad hard drive or is it a bad flex cable pretending to be a bad hard drive? There must be some way to see/test/diagnose sata errors to verify a bad cable?

 

[l008com]

Are you the first person to see these Macbooks? Could the previous tech who worked on them have damaged the cable?

And yes, Computer with bad cables do crash and they do freeze - at least computers running a UNIX/Linux operating system. UNIX/Linux are far less tolleratnt to losing their disk than windows. I know from a VMWare/ESXi perspective, when we would have a NFS datastore outage, the Linux VMs would kernel panic where the Windows VMs would start working again once the datastore appeared and would act as if nothing happened. We had to increase the timeout policy for the kernel panic to allow the Linux VMs to survive outages for a specific period. If that time period threshold was breached, the VM would panic.

Yes I am the first person, and usually original drives.
And the issue with these bad cables isn't that the drive disappears. It's still there, it's just random read/write calls seem to hang. It's exactly the same symptom as a hard drive with bad blocks.

 

[Archer1212]

But if you're eliminating the flex cable by booting over firewire or ethernet, you are also eliminating the hard drive. Which leads you right back to the problem of: is it really a bad hard drive or is it a bad flex cable pretending to be a bad hard drive? There must be some way to see/test/diagnose sata errors to verify a bad cable?

How is booting over firewire/usb eliminating the hard drive? the drive is what we were booting to, just using a slower connection.

 

[l008com]

How is booting over firewire/usb eliminating the hard drive? the drive is what we were booting to, just using a slower connection.

You mean put the original drive in a firewire enclosure, and boot the original drive?

Well ok that would work but that leaves me right where I am when I started. That leaves me with what seems like a bad flex cable, but not confirmed for sure. It's the same as if I had left the hard drive in, but ran SMART Utility to verify that the drive has no issues. But I'm still experiencing beachballs as if the drive were failing. That strongly implies that the cable is bad, just like booting the original hard drive in an enclosure would. But I want to confirm a bad cable. I want something like a console full of sata bus errors.

 

[Archer1212]

You mean put the original drive in a firewire enclosure, and boot the original drive?

Well ok that would work but that leaves me right where I am when I started. That leaves me with what seems like a bad flex cable, but not confirmed for sure. It's the same as if I had left the hard drive in, but ran SMART Utility to verify that the drive has no issues. But I'm still experiencing beachballs as if the drive were failing. That strongly implies that the cable is bad, just like booting the original hard drive in an enclosure would. But I want to confirm a bad cable. I want something like a console full of sata bus errors.

How is that not confirming the cable? That is the only piece of actual hardware you are bypassing. You can then run any hard drive tests afterwards. If you are still having performance issues with the drive in an enclosure then you know your issue is not the cable. Could be the drive, OS, logicboard, or third party software.
Apple has no way to confirm the cable is bad. Its a cable, it doesnt have some sort of co-processor that saves logs and reports back to the OS. You bypass the cable and if the issue is resolved then its the cable. If the Mac OS saves any errors it is going to be in the Console logs, but from my memory it doesnt save any logs related to that.