smartpqi bug / driver update? HPE System stuck with "bus_dmamap_load_ccb failed = 36 count = 131072"

[nielsd]

Hi Dears,


i try to install current stable TrueNAS 12 on a HP DL360 Gen 10 with
Intel(R) Xeon(R) Silver 4208 CPU @ 2.10GHz and
HPE Smart Array P408i-a SR Gen10 HBA (with latest firmware 4.11)
4 x 2.4 TB HPE disks (with firmware HPD6)

The HBA is "unconfigured" (means HBA mode?) - all attached disks are used in ZFS

Installation works well, but if i try to do some interaction like installing packages in jails, console throws errors and the process stucks unkillable - this renders the whole system instable/unsusable after that. Just a hard reset then works.

ERROR:

Code:

[CPU 7][pqi_map_request][540]:bus_dmamap_load_ccb failed = 36 count = 131072

beside this, i can see several lines in dmesg from time to time (smartd is disabled)

ERROR:

Code:

[CPU 2][os_aio_response_error] SCSI_STATUS_CHECK_COND sense size 36

to what i've found in the FreeBSD space, the error seems related to the smartpqi driver - it seems the bug is related to the FreeBSD smartpqi driver:

[Bug 257890] Storage controller lockup on zfs scrub [smartpqi][zfs]

www.mail-archive.com

240145 – [smartpqi][zfs] kernel panic with hanging vdev

bugs.freebsd.org

⚙ D30182 Newly added features and bug fixes in latest Microchip SmartPQI driver for FreeBSD13.0

reviews.freebsd.org

is there any chance to get a current driver backported to TrueNAS 12 or - if not - in the upcoming 13 version?
Or does anyone any workaround so far?

If i'm right, this should be this:

Adaptec by PMC | Page Not Found

Many thanks for your time,


niels.

 

[Ericloewe]

It’s interesting that Microchip seems to be participating, for a change. Adaptec has a well-earned reputation for being damn near unusable in FreeBSD.

 

[nielsd]

Just to follow up:
The bug is solved within the newer FreeBSD kernels of TrueNAS-13.0-U2.x or TrueNAS-13.0-U3.x

hth,
cheers,

niels.

 

[jgreco]

Just to follow up:
The bug is solved within the newer FreeBSD kernels of TrueNAS-13.0-U2.x or TrueNAS-13.0-U3.x

hth,
cheers,

niels.

Regardless, you need to be aware that this is essentially an unknown driver that may be unreliable in a variety of ways. Please see the article

What's all the noise about HBAs, and why can't I use a RAID controller?

1) An HBA is a Host Bus Adapter. This is a controller that allows SAS and SATA devices to be attached to, and communicate directly with, a server. RAID controllers typically aggregate several disks into a Virtual Disk abstraction of some sort...

www.truenas.com

It's possible that this new driver could make these controllers usable, but it's relatively unlikely they got all the details right. If you want to try it, be aware that you're being a guinea pig for Adaptec. Have a good backup of everything on your pool.

 

[Ericloewe]

Do let us know how it goes, either way.

 

[jgreco]

Adaptec has a well-earned reputation for being damn near unusable in FreeBSD.

That, incidentally, is a great tragedy. The Adaptec SCSI controllers like the AHA-1542CF were the best in breed back in the day. Still have some in inventory somewhere.