HP H240 Smart HBA Support (in HBA Mode)


ezhik

Recommended Posts

The 6.3 release added the hpsa (HP Smart Array) module, which added support for your H240, just what you needed.  But it doesn't appear to be working correctly:

 

* The HP card was recognized and initialized, looks correct, and it's 7 drives were identified and initialized.  Their partition tables were read in fine.  You started the array in Maintenance mode, quickly stopped it, then started it in regular mode.  The BTRFS file systems of the data drives and the Cache Pool (a pair of Patriot SSD's) were all loaded without issue.  But only half a minute after array start, the H240 declares a fault and aborts, and many of the connected drives are declared offline, which appears to be fatal for this session.  unRAID is not aware the card has stopped (or almost stopped), and tries to read from the drives, which isn't possible now, so numerous read errors are reported instead.  Nothing actually wrong with the drives, but they can't be accessed because the card appears to have crashed.  Finally BTRFS crashes too when it can't read from the drives, causing the kernel issues.

 

It's hard to make any conclusions yet.  You appear to be the first to test hpsa support, so we cannot say whether it's compatible or not.  It was added because it was requested.  And because a small company can't possibly test every card, it has to rely often on user reports as to what works and what doesn't.  It could be a defective card, it could be a defective driver, the card could be incompatible with your motherboard, the driver may be too new and not stable enough yet, or it may not be configured correctly.  Look for a firmware update for the card, and see if that helps.

 

You would be MUCH better off going with tried and tested hardware, already known to be compatible!

 

* Check for a newer BIOS for your motherboard, might help.

 

* You have IDE emulation turned on for your onboard SATA drives.  When you next boot, go into the BIOS settings and look for the SATA mode, and change it to a native SATA mode, preferably AHCI, anything but IDE emulation mode.  It should be slightly faster, and a little safer.  You aren't using any of your 4 motherboard SATA ports, but once configured for AHCI, they are your best SATA ports.  (The H240 may be better for high performance SSD's though.)

Link to comment

The 6.3 release added the hpsa (HP Smart Array) module, which added support for your H240, just what you needed.  But it doesn't appear to be working correctly:

 

* The HP card was recognized and initialized, looks correct, and it's 7 drives were identified and initialized.  Their partition tables were read in fine.  You started the array in Maintenance mode, quickly stopped it, then started it in regular mode.  The BTRFS file systems of the data drives and the Cache Pool (a pair of Patriot SSD's) were all loaded without issue.  But only half a minute after array start, the H240 declares a fault and aborts, and many of the connected drives are declared offline, which appears to be fatal for this session.  unRAID is not aware the card has stopped (or almost stopped), and tries to read from the drives, which isn't possible now, so numerous read errors are reported instead.  Nothing actually wrong with the drives, but they can't be accessed because the card appears to have crashed.  Finally BTRFS crashes too when it can't read from the drives, causing the kernel issues.

 

It's hard to make any conclusions yet.  You appear to be the first to test hpsa support, so we cannot say whether it's compatible or not.  It was added because it was requested.  And because a small company can't possibly test every card, it has to rely often on user reports as to what works and what doesn't.  It could be a defective card, it could be a defective driver, the card could be incompatible with your motherboard, the driver may be too new and not stable enough yet, or it may not be configured correctly.  Look for a firmware update for the card, and see if that helps.

 

You would be MUCH better off going with tried and tested hardware, already known to be compatible!

 

* Check for a newer BIOS for your motherboard, might help.

 

* You have IDE emulation turned on for your onboard SATA drives.  When you next boot, go into the BIOS settings and look for the SATA mode, and change it to a native SATA mode, preferably AHCI, anything but IDE emulation mode.  It should be slightly faster, and a little safer.  You aren't using any of your 4 motherboard SATA ports, but once configured for AHCI, they are your best SATA ports.  (The H240 may be better for high performance SSD's though.)

 

Rob you are correct.

 

The fault is declared when I am attempting to copy the data from the array (read it).

 

I changed the SATA mode to AHCI. It made no difference.

 

Are you using one of these drivers?

 

http://h20564.www2.hpe.com/hpsc/swd/public/readIndex?sp4ts.oid=7553524&swLangOid=8&swEnvOid=4049

 

It's running latest 4.52 firmware.

 

I am willing to do as much testing as possible on this. I would really appreciate it if we could make this working.

 

Cheers.

 

If sending you a card would be better, I have no problem with sending one.

Link to comment

I changed the SATA mode to AHCI. It made no difference.

The BIOS SATA mode option only applies to the motherboard ports, which you aren't using.  It will make a difference whenever you decide to use them.

 

Are you using one of these drivers?

I'm a user like you, except I can't afford anything that expensive.  You and whoever first requested it are probably the only users, so you'll probably have to do your own testing.  I doubt anyone else, including anyone at Lime Technology, has one.

 

The card almost works, in that it appeared to go as far as mounting the drives and their file systems without apparent issue.  So it's not an IT mode issue, or any other known issue we've seen.  You're sure the card does work, has worked before in other applications?  Make sure it's seated fine, is kept sufficiently cool, and has sufficient power.  Try moving 4 drives off it, to the motherboard, and see if it can handle a lighter load.

Link to comment

So I've tested this on Ubuntu 16.04. Works without any issues. Modules loaded: hpsa and cciss.

 

$ pv ubuntu-gnome-16.04.1-desktop-amd64.iso > ~/Downloads/ubuntu-gnome-16.04.1-desktop-amd64.iso

1.21GiB 0:00:01 [ 631MiB/s] [====================================================================================>] 100%

 

# modinfo hpsa

filename:      /lib/modules/4.4.0-47-generic/kernel/drivers/scsi/hpsa.ko

license:        GPL

version:        3.4.14-0

description:    Driver for HP Smart Array Controller version 3.4.14-0

author:        Hewlett-Packard Company

srcversion:    EFB2AD999EE7878A33900C5

depends:        scsi_transport_sas

intree:        Y

vermagic:      4.4.0-47-generic SMP mod_unload modversions

parm:          hpsa_allow_any:Allow hpsa driver to access unknown HP Smart Array hardware (int)

parm:          hpsa_simple_mode:Use 'simple mode' rather than 'performant mode' (int)

#

 

# modinfo cciss

filename:      /lib/modules/4.4.0-47-generic/kernel/drivers/block/cciss.ko

license:        GPL

version:        3.6.26

description:    Driver for HP Smart Array Controllers

author:        Hewlett-Packard Company

srcversion:    07CA5DF6C8E8F4A1C5D384D

depends:       

intree:        Y

vermagic:      4.4.0-47-generic SMP mod_unload modversions

parm:          cciss_tape_cmds:number of commands to allocate for tape devices (default: 6) (int)

parm:          cciss_simple_mode:Use 'simple mode' rather than 'performant mode' (int)

parm:          cciss_allow_hpsa:Prevent cciss driver from accessing hardware known to be  supported by the hpsa driver (int)

#

 

 

Link to comment
  • 1 month later...
  • 4 weeks later...
  • 1 month later...

What is the issue you are facing?

 

I quickly tested this on X8SIL-F with Xeon L3426.

 

The card and the drives got detected by booting a *clean* 6.3.2. Not an upgrade, I did a clean system test (trial).

 

The drives were detected and it started to pre-clear the drives without any errors.

 

Mind you, my upgrades did not work as intended. Every time I'd need to write to the drives, it would crash.

 

But a clean setup seems to be working.

Link to comment

I can't see any of the disks in unraid, I can fine see it in windows and i can see disks connected to the motherboard, but not any of the disks connected to the controller.

 

I tried to use a clean installation, even on a new USB to be sure.

 

If i swap to my D3116C RAID controller and make every disk in a raid 0, I can see each disk, but this is not a working solution, as I lose all monitoring on the disks and it will be a problem to recover if the controller fails.

 

I have attached the log from the system, it have been analysed before, but don't know how to fix that problem

tower-diagnostics-20170323-1010.zip

 

https://forums.lime-technology.com/topic/55974-cant-see-hp-h240-hba/#comment-548071

Link to comment

Just tried to put back the H240 after I tested the D3116C, first time i booted a ESXI 6.5 with HPE Drivers i see all disks, if I then reboot all disks are gone again, but no mater what time I boot to Windows Server 2012 R2 I can always see the disks?

Link to comment
7 hours ago, Tvede said:

I can't see any of the disks in unraid, I can fine see it in windows and i can see disks connected to the motherboard, but not any of the disks connected to the controller.

 

I tried to use a clean installation, even on a new USB to be sure.

 

If i swap to my D3116C RAID controller and make every disk in a raid 0, I can see each disk, but this is not a working solution, as I lose all monitoring on the disks and it will be a problem to recover if the controller fails.

 

The card or driver is broken, not working.  Here are the relevant syslog sections:

Quote

 

Mar 23 10:08:37 Tower kernel: pci_bus 0000:01: resource 1 [mem 0xf7b00000-0xf7bfffff]

...

Mar 23 10:08:37 Tower kernel: pci 0000:01:00.0: can't claim BAR 6 [mem 0xf7b04000-0xf7b83fff pref]: address conflict with 0000:01:00.0 [mem 0xf7b0c000-0xf7b0c3ff 64bit]
Mar 23 10:08:37 Tower kernel: pci 0000:01:00.0: BAR 0: no space for [mem size 0x00100000 64bit]
Mar 23 10:08:37 Tower kernel: pci 0000:01:00.0: BAR 0: trying firmware assignment [mem 0xf7b0d000-0xf7c0cfff 64bit]
Mar 23 10:08:37 Tower kernel: pci 0000:01:00.0: BAR 0: [mem 0xf7b0d000-0xf7c0cfff 64bit] conflicts with PCI Bus 0000:00 [mem 0xdf200000-0xfeafffff window]
Mar 23 10:08:37 Tower kernel: pci 0000:01:00.0: BAR 0: failed to assign [mem size 0x00100000 64bit]
Mar 23 10:08:37 Tower kernel: pci 0000:01:00.0: BAR 6: assigned [mem 0xf7b80000-0xf7bfffff pref]

...

Mar 23 10:08:37 Tower kernel: HP HPSA Driver (v 3.4.16-0)
Mar 23 10:08:37 Tower kernel: hpsa 0000:01:00.0: can't disable ASPM; OS doesn't have ASPM control
Mar 23 10:08:37 Tower kernel: hpsa 0000:01:00.0: MSI-X capable controller
Mar 23 10:08:37 Tower kernel: hpsa 0000:01:00.0: MSI-X init failed -12
Mar 23 10:08:37 Tower kernel: hpsa 0000:01:00.0: MSI capable controller

...

Mar 23 10:08:40 Tower kernel: hpsa 0000:01:00.0: board not ready, timed out.

 

 

Looks like memory region conflicts (possibly with itself?!?).  Both Microsoft and Linux kernel devs have gotten good about detecting and working around hardware 'quirks', and perhaps Microsoft has done a better job here.  You can try reconfiguring the card, to see if it will improve its PCI declarations.  Or perhaps Tom ( @limetech ) will have an idea, once he sees this.

 

Could also be useful to see @ezhik's syslog, to compare the same sections.

Link to comment
5 hours ago, Tvede said:

What do you mean by reconfigure? I have reset the card, removed disks and so on?

 

I don't know, for a specific card.  But some cards have their own BIOS setup screen, or jumpers, a way to make significant changes in the PCI configuration of the card.  I have no idea what your card has.

 

5 hours ago, Tvede said:

Are unraid using a "up to date" driver? maybe there is a new one with fixes

 

Perhaps Tom will check on that, can't speak for him.  But remember, it's not clear what the issue is, may not be the driver.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.