HDD event ID 153 errors, Windows 10 and AHCI Link Power Management

[magellan]

I was getting a lot of HDD event ID 153 errors on some older HDD's I have (an ancient, refurbished WD 1600js [that has both a 4-pin molex power connector and SATA power] and a Toshiba HDWD120) but never on my pair of Samsung 850 Pro 256GiB SATA SSD's. There's an interesting hidden setting called AHCI Link power management under advanced power settings in power options under hard disk for HIPM/DIPM that defaults to HIPM (Host Initiated Power Management) I believe. You have to change the registry to enable access to this setting in Windows 10. This controls the power state of the SATA link, powering it down when not in use. I'm thinking both my HDD's are too old to properly recognize AHCI link power management and this was why I was getting the event ID 153 errors. This setting is also dependent on BIOS settings as well (e.g. SATA Aggressive link power management) and having the HDD controller set to ACHI mode (not RAID).

So if I were to change the AHCI Link Power Management to DIPM (Device Initiated Power Management), would just the SSD's power down their SATA connections? Or would the SATA AHCI HDD controller respond to the DIPM request and power down ALL SATA links (incl. those to my HDD's) instead of just the SSD's?

 

[EarthDog]

I mean, if that is your guess, I'm sure you can look up if the drives support it, right? Looks like your 160GB drive is SATA I based and doesn't support AHCI...however your 2TB Toshiba is SATA III and supports AHCI. What system are these drives on? Will you please update your signature so we know what hardware you're working with, exactly?

That said, a 153 is this:

This event may be an indication that the specified drive may be failing soon. This is especially true if the event is recorded regularly. The first steps is to run the maintenance utilities related to hard drives, starting with CHKDSK and the utilities available from the manufacturer for that specific brand.

It is recommended to ensure that the data on that drive is on a backup schedule. The drive should be replaced as soon as possible.

http://www.eventid.net/display-eventid-153-source-disk-eventno-11182-phase-1.htm

A system administrator who encounters event 153 errors should investigate the health of the computerÂ’s disk subsystem. Although an occasional timeout may be part of the normal operation of a system, the frequent need to retry requests indicates a performance issue with the storage that should be corrected.

https://docs.microsoft.com/en-us/archive/blogs/ntdebugging/interpreting-event-153-errors

I don't see anywhere in those links that has to do with power. This is not something I would mess with. Can you please post an image showing that error and any details?

Have you run chkdsk against it? What about any other drive diagnostic tool and what does it say? You say other drives don't get it, buuuuuuuuuut, they are new(er) SATA based SSDs. Maybe your self-proclaimed 'ancient' drives are about to crap out....

 

[magellan]

SMART doesn't give any indication my HDD's are failing, Properties --> Tools --> Check came up with no errors found on the WD 160GiB HDD and all the partitions on the Toshiba 2 TiB ("windows successfully scanned the drive. No errors were found"). Since I modified the power profile to AHCI Link Power Management to DIPM two days ago I haven't had any event ID 153's show up in my log yet. I'd figure just because a HDD is AHCI compliant doesn't necessarily mean it complies with AHCI Link Power Management protocols.

I updated my profile with my new system, but that doesn't show up in my signature, so I'll have to change that too.

Here's an example of one such event ID 153:

The IO operation at logical block address 0x14f840 for Disk 2 (PDO name: \Device\00000034) was retried.

- EventData

\Device\Harddisk2\DR2
0x14f840
2
\Device\00000034
0F01040004002C00000000009900048000000000000000000000000000000000000000000000000000000428

the last three bytes indicate:
the SCSI status is 00 (good),
the SRB status is 04 (the request completed with any other error),
and the command was 28 (read) and not 2A (write).

 

[EarthDog]

Yes, update your signature, not profile.

What about HD Tune? Passmark? None of those read any bad sectors or anything?

I modified the power profile to AHCI Link Power Management to DIPM two days ago I haven't had any event ID 153's show up in my log yet. I'd figure just because a HDD is AHCI compliant doesn't necessarily mean it complies with AHCI Link Power Management protocols.

Wait.........so, are you telling us something, or asking for help? I'm confused because the first post appears like you're asking for help, but this last post infers you tried this days ago and it has fixed the problem so far. That would be great information to have out of the gate, dude ...........................

 

[magellan]

I'm not sure the problem is fixed, it's only been a couple of days and the event ID 153's don't happen all the time.

 

[EarthDog]

Awesome man.. hope that was it. Don't forget to report back!

 

[magellan]

I have a HP dvd1260i DVD Writer that used to be installed internally with my Asrock z390 Taichi but would regularly cease working, so I figured it was dead. On a lark I installed it in my x79 Asrock Extreme 4 rig though and it worked perfectly. I also had regular Event ID 153: disk that seem to have cleared up once I set power options --> advanced options --> hard disk --> AHCI Link Power Management to DIPM (from the default of HIPM). I believe HIPM means the host (i.e. SATA controller) will initiate SATA link power down even if the device doesn't support it. If my DVD-RW didn't support AHCI link power management could this have been why it would always cease functioning on the Asrock z390 Taichi once the OS booted? One other question: can the BIOS itself initiate AHCI link power management all on its own or does it have to be initiated by an OS?

 

[EarthDog]

Merged like threads. This seems alllllllllllllll about power management to HDD/DVD etc.

EDIT: To answer your first question, X79 had AHCI in it too... unless you had it disabled and running IDE mode on it... Assuming it was enabled on both, then no, that wouldn't be the reason.

Pretty sure the OS has to support it. IIRC, Vista was the first OS where AHCI was introduced.

 

[magellan]

Merged like threads. This seems alllllllllllllll about power management to HDD/DVD etc.

EDIT: To answer your first question, X79 had AHCI in it too... unless you had it disabled and running IDE mode on it... Assuming it was enabled on both, then no, that wouldn't be the reason.

Pretty sure the OS has to support it. IIRC, Vista was the first OS where AHCI was introduced.

Thanks Earthdog, the x79 rig is still running Windows XP. It's hard to imagine power management settings causing these kinds of hardware failures.

 

[EarthDog]

It's hard to imagine power management settings causing these kinds of hardware failures

Agreed... just answering what's asked.