HP Proliant dl380 g6

[Carpesimia]

Hello,

 

I got my hands on the above server to create an unraid machine.  My drives and caddies have arrived, so I decided to set it up.

 

I created the USB drive (and approved one on the list) as instructed, and set up the CMOS for all settings.  I am very much able to boot into unraid.

 

The issue is, it seems when it comes to sensing my drives.  Even with the original HP drives that came with the server, unraid does not see them.  Here's some stuff:

 

from SYSTEM DEVICES:

 

SCSI Devices

[0:0:0:0]    disk    SanDisk  Cruzer Fit       1.00  /dev/sda   15.6GB
[1:1:0:0]    disk    HP       LOGICAL VOLUME   1.66  /dev/sdb   1.00TB
[1:3:0:0]    storage HP       P410i            1.66  -               -
[2:0:0:0]    cd/dvd  TEAC     DV-W28S-VS       G.V0  /dev/sr0        -

 

I only have one of my drives in now, but unraid can clearly see the 1TB drive.  But, over in the syslog I see this in one spot:

 

ec 31 13:00:20 Proliant kernel: scsi host1: hpsa
Dec 31 13:00:20 Proliant kernel: hpsa 0000:04:00.0: scsi 1:0:0:0: masked Direct-Access     ATA      ST91000640NS     PHYS DRV SSDSmartPathCap- En- Exp=0
Dec 31 13:00:20 Proliant kernel: hpsa 0000:04:00.0: scsi 1:0:1:0: masked Enclosure         PMCSIERA  SRC 8x6G        enclosure SSDSmartPathCap- En- Exp=0
Dec 31 13:00:20 Proliant kernel: hpsa 0000:04:00.0: scsi 1:1:0:0: added Direct-Access     HP       LOGICAL VOLUME   RAID-0 SSDSmartPathCap- En- Exp=1
Dec 31 13:00:20 Proliant kernel: hpsa 0000:04:00.0: scsi 1:3:0:0: added RAID              HP       P410i            controller SSDSmartPathCap- En- Exp=1
Dec 31 13:00:20 Proliant kernel: scsi 1:1:0:0: Direct-Access     HP       LOGICAL VOLUME   1.66 PQ: 0 ANSI: 5
Dec 31 13:00:20 Proliant kernel: sd 1:1:0:0: [sdb] 1953459632 512-byte logical blocks: (1.00 TB/931 GiB)
Dec 31 13:00:20 Proliant kernel: sd 1:1:0:0: Attached scsi generic sg1 type 0
Dec 31 13:00:20 Proliant kernel: sd 1:1:0:0: [sdb] Write Protect is off
Dec 31 13:00:20 Proliant kernel: scsi 1:3:0:0: RAID              HP       P410i            1.66 PQ: 0 ANSI: 0
Dec 31 13:00:20 Proliant kernel: sd 1:1:0:0: [sdb] Mode Sense: 6b 00 00 08
Dec 31 13:00:20 Proliant kernel: sd 1:1:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
Dec 31 13:00:20 Proliant kernel: scsi 1:3:0:0: Attached scsi generic sg2 type 12
Dec 31 13:00:20 Proliant kernel: sd 1:1:0:0: [sdb] Attached SCSI disk
Dec 31 13:00:20 Proliant kernel: DMAR: 32bit 0000:00:1f.2 uses non-identity mapping
Dec 31 13:00:20 Proliant kernel: scsi host2: ata_piix
Dec 31 13:00:20 Proliant kernel: scsi host3: ata_piix
Dec 31 13:00:20 Proliant kernel: ata1: SATA max UDMA/133 cmd 0x1080 ctl 0x1088 bmdma 0x10a0 irq 17
Dec 31 13:00:20 Proliant kernel: ata2: SATA max UDMA/133 cmd 0x1090 ctl 0x1098 bmdma 0x10a8 irq 17
Dec 31 13:00:20 Proliant kernel: ipmi_si 0000:01:04.6: Using irq 21
Dec 31 13:00:20 Proliant kernel: ipmi_si 0000:01:04.6: Found new BMC (man_id: 0x00000b, prod_id: 0x2000, dev_id: 0x11)
Dec 31 13:00:20 Proliant kernel: ipmi_si 0000:01:04.6: IPMI kcs interface initialized
Dec 31 13:00:20 Proliant kernel: ata2.00: SATA link down (SStatus 4 SControl 300)
Dec 31 13:00:20 Proliant kernel: ata2.01: SATA link down (SStatus 4 SControl 300)
Dec 31 13:00:20 Proliant kernel: ata1.00: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
Dec 31 13:00:20 Proliant kernel: ata1.01: SATA link down (SStatus 4 SControl 300)
Dec 31 13:00:20 Proliant kernel: ata1.01: link offline, clearing class 3 to NONE
Dec 31 13:00:20 Proliant kernel: ata1.00: ATAPI: TEAC    DV-W28S-VS, , G.V0, max UDMA/100
Dec 31 13:00:20 Proliant kernel: ata1.00: configured for UDMA/100
Dec 31 13:00:20 Proliant kernel: scsi 2:0:0:0: CD-ROM            TEAC     DV-W28S-VS       G.V0 PQ: 0 ANSI: 5
Dec 31 13:00:20 Proliant kernel: sr 2:0:0:0: [sr0] scsi3-mmc drive: 24x/24x writer dvd-ram cd/rw xa/form2 cdda tray
Dec 31 13:00:20 Proliant kernel: cdrom: Uniform CD-ROM driver Revision: 3.20
Dec 31 13:00:20 Proliant kernel: sr 2:0:0:0: Attached scsi CD-ROM sr0
Dec 31 13:00:20 Proliant kernel: sr 2:0:0:0: Attached scsi generic sg3 type 5
Dec 31 13:00:20 Proliant kernel: NMI: PCI system error (SERR) for reason a1 on CPU 0.
Dec 31 13:00:20 Proliant kernel: Dazed and confused, but trying to continue
Dec 31 13:00:20 Proliant kernel: DMAR: DRHD: handling fault status reg 2
Dec 31 13:00:20 Proliant kernel: DMAR: [DMA Read] Request device [04:00.0] fault addr ffe90000 [fault reason 06] PTE Read access is not set
Dec 31 13:00:20 Proliant kernel: random: crng init done

 

And then further down:

Dec 31 13:00:23 Proliant emhttp: shcmd (6): modprobe md-mod super=/boot/config/super.dat |& logger
Dec 31 13:00:23 Proliant kernel: md: unRAID driver 2.7.2 installed
Dec 31 13:00:24 Proliant emhttp: Trial key detected, GUID: 0781-5571-0001-291031116055 FILE: /boot/config/Trial.key
Dec 31 13:00:24 Proliant emhttp: Device inventory:
Dec 31 13:00:24 Proliant emhttp: shcmd (7): udevadm settle
Dec 31 13:00:32 Proliant kernel: hpsa 0000:04:00.0: scsi 1:1:0:0: resetting logical  Direct-Access     HP       LOGICAL VOLUME   RAID-0 SSDSmartPathCap- En- Exp=1
Dec 31 13:00:38 Proliant login[1967]: ROOT LOGIN  on '/dev/tty1'
Dec 31 13:01:02 Proliant emhttp: err: ckmbr: open: /dev/sdb No such device or address
Dec 31 13:01:02 Proliant emhttp: device /dev/sdb problem getting size
Dec 31 13:01:02 Proliant kernel: hpsa 0000:04:00.0: Command timed out.
Dec 31 13:01:02 Proliant kernel: hpsa 0000:04:00.0: Failed to send reset command
Dec 31 13:01:02 Proliant kernel: hpsa 0000:04:00.0: scsi 1:1:0:0: reset logical  failed Direct-Access     HP       LOGICAL VOLUME   RAID-0 SSDSmartPathCap- En- Exp=1
Dec 31 13:01:02 Proliant kernel: sd 1:1:0:0: Device offlined - not ready after error recovery

 

And after that, I see this repeated several times:

Dec 31 13:01:22 Proliant emhttp: import 30 cache device: no device
Dec 31 13:01:22 Proliant emhttp: import flash device: sda
Dec 31 13:01:23 Proliant emhttp: shcmd (43): rmmod md-mod |& logger
Dec 31 13:01:23 Proliant kernel: md: unRAID driver removed
Dec 31 13:01:23 Proliant emhttp: shcmd (44): modprobe md-mod super=/boot/config/super.dat |& logger
Dec 31 13:01:23 Proliant kernel: md: unRAID driver 2.7.2 installed
Dec 31 13:01:23 Proliant emhttp: Trial key detected, GUID: 0781-5571-0001-291031116055 FILE: /boot/config/Trial.key
Dec 31 13:01:23 Proliant emhttp: Device inventory:
Dec 31 13:01:23 Proliant emhttp: shcmd (45): udevadm settle
Dec 31 13:01:23 Proliant emhttp: err: ckmbr: open: /dev/sdb No such device or address
Dec 31 13:01:23 Proliant emhttp: device /dev/sdb problem getting size
Dec 31 13:01:23 Proliant emhttp: SanDisk_Cruzer_Fit_4C530001291031116055-0:0 (sda) 15297504

 

Does this tell anyone anything?  I see it being recognized in the system devices, but then it doesnt show up as a selectable device.

 

And I posting this in the right spot?  I havent purchased unraid (yet), because I wanted to get this up and working first.

 

Thanks!

 

 

[1812]

see the link in my sig for proliant tips/tricks/mods you need to do.

 

unRaid can be finicky with p410i controllers at times. More so when it is used in conjunction with another disk accessing controller (like onboard sata or others) The p410i does not have JBOD setting so you have to make a raid 0 out of each disk to pass it to unRaid, otherwise it will never see them.

 

If you're running unRaid version 6.3.5 either downgrade to 6.2.4 or move to the 6.4 release client ( I moved all 3 of my hp servers to that already.) 

 

 

[Carpesimia]

First, thank you for that response.   While I admit to being technical, I am new to the Proliant and to unraid, so please entertain my questions with a grain of salt.

 

So, if I understand right, there's two things I'm going to need to do:

 

1) Download the 6.2.4 version of unraid on my USB, instead of the one I got.

 

2) Add each drive to it's own raid-0 raid (if I can figure that part out.. lol).

 

But that does leave me with a question.  You mentioned in the other thread that if you intend to use the p410 controller.  Do I have a choice?  With those hot-swap drive bays, dont they all directly connect to the 410 with no alternative?

 

I'll work on the other things in the meantime.  Sorry for the newbish questions, and thank's for the quick response.

 

 

[Carpesimia]

Ok, so updates:

 

1) Downloaded 6.2.4 on another USB (I bought 2), and made it bootable.  Swapped it out in the proliant.

 

2) Went into the proliant raid utility, where i found that the one drive that I have installed right now is already configured as a 1-drive raid-0.   In any case, I went ahead and deleted the logical drive and re-created it, again as a raid 0.

 

3) Booted the system, unraid came up.  Verified my IP then went to my other PC.

 

4) Opened unraid in the browser. Verified version 6.2.4, and looked in system devices, and it looked correct:

 

SCSI Devices

[0:0:0:0]    disk    SanDisk  Cruzer Fit       1.00  /dev/sda   15.6GB
[1:1:0:0]    disk    HP       LOGICAL VOLUME   1.66  /dev/sdb   1.00TB
[1:3:0:0]    storage HP       P410i            1.66  -               -
[2:0:0:0]    cd/dvd  TEAC     DV-W28S-VS       G.V0  /dev/sr0        -

 

Then, I looked in main to select a drive.  Again, the drop-downs are all empty.  So back to the system log.  Here are some of the entries I see that could be related:

 

Dec 31 11:36:23 Tower kernel: scsi host1: hpsa
Dec 31 11:36:23 Tower kernel: hpsa 0000:04:00.0: scsi 1:0:0:0: masked Direct-Access     ATA      ST91000640NS     PHYS DRV SSDSmartPathCap- En- Exp=0
Dec 31 11:36:23 Tower kernel: hpsa 0000:04:00.0: scsi 1:0:1:0: masked Enclosure         PMCSIERA  SRC 8x6G        enclosure SSDSmartPathCap- En- Exp=0
Dec 31 11:36:23 Tower kernel: hpsa 0000:04:00.0: scsi 1:1:0:0: added Direct-Access     HP       LOGICAL VOLUME   RAID-0 SSDSmartPathCap- En- Exp=1
Dec 31 11:36:23 Tower kernel: hpsa 0000:04:00.0: scsi 1:3:0:0: added RAID              HP       P410i            controller SSDSmartPathCap- En- Exp=1
Dec 31 11:36:23 Tower kernel: scsi 1:1:0:0: Direct-Access     HP       LOGICAL VOLUME   1.66 PQ: 0 ANSI: 5
Dec 31 11:36:23 Tower kernel: sd 1:1:0:0: Attached scsi generic sg1 type 0
Dec 31 11:36:23 Tower kernel: sd 1:1:0:0: [sdb] 1953459632 512-byte logical blocks: (1.00 TB/931 GiB)
Dec 31 11:36:23 Tower kernel: sd 1:1:0:0: [sdb] Write Protect is off
Dec 31 11:36:23 Tower kernel: sd 1:1:0:0: [sdb] Mode Sense: 6b 00 00 08
Dec 31 11:36:23 Tower kernel: sd 1:1:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
Dec 31 11:36:23 Tower kernel: scsi 1:3:0:0: RAID              HP       P410i            1.66 PQ: 0 ANSI: 0
Dec 31 11:36:23 Tower kernel: scsi 1:3:0:0: Attached scsi generic sg2 type 12
Dec 31 11:36:23 Tower kernel: sd 1:1:0:0: [sdb] Attached SCSI disk
Dec 31 11:36:23 Tower kernel: DMAR: 32bit 0000:00:1f.2 uses non-identity mapping
Dec 31 11:36:23 Tower kernel: scsi host2: ata_piix
Dec 31 11:36:23 Tower kernel: scsi host3: ata_piix
Dec 31 11:36:23 Tower kernel: ata1: SATA max UDMA/133 cmd 0x1080 ctl 0x1088 bmdma 0x10a0 irq 17
Dec 31 11:36:23 Tower kernel: ata2: SATA max UDMA/133 cmd 0x1090 ctl 0x1098 bmdma 0x10a8 irq 17
Dec 31 11:36:23 Tower kernel: ipmi_si 0000:01:04.6: Using irq 21
Dec 31 11:36:23 Tower kernel: ipmi_si 0000:01:04.6: Found new BMC (man_id: 0x00000b, prod_id: 0x2000, dev_id: 0x11)
Dec 31 11:36:23 Tower kernel: ipmi_si 0000:01:04.6: IPMI kcs interface initialized
Dec 31 11:36:23 Tower kernel: ata2.00: SATA link down (SStatus 4 SControl 300)
Dec 31 11:36:23 Tower kernel: ata2.01: SATA link down (SStatus 4 SControl 300)
Dec 31 11:36:23 Tower kernel: ata1.00: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
Dec 31 11:36:23 Tower kernel: ata1.01: SATA link down (SStatus 4 SControl 300)
Dec 31 11:36:23 Tower kernel: ata1.01: link offline, clearing class 3 to NONE
Dec 31 11:36:23 Tower kernel: ata1.00: ATAPI: TEAC    DV-W28S-VS, , G.V0, max UDMA/100
Dec 31 11:36:23 Tower kernel: ata1.00: configured for UDMA/100
Dec 31 11:36:23 Tower kernel: scsi 2:0:0:0: CD-ROM            TEAC     DV-W28S-VS       G.V0 PQ: 0 ANSI: 5
Dec 31 11:36:23 Tower kernel: sr 2:0:0:0: [sr0] scsi3-mmc drive: 24x/24x writer dvd-ram cd/rw xa/form2 cdda tray

 

So the above sees the disk, and the size and attaches the disk.  Then I see a few link downs, but one up.  I'm a little disappointed it says 1.5 Gbps, but at this point I just want to see it work.  Moving on.

Dec 31 11:36:29 Tower kernel: md: unRAID driver 2.6.8 installed
Dec 31 11:36:29 Tower kernel: read_file: error 2 opening /boot/config/super.dat
Dec 31 11:36:29 Tower kernel: md: could not read superblock from /boot/config/super.dat
Dec 31 11:36:29 Tower kernel: md: initializing superblock
Dec 31 11:36:29 Tower emhttp: Unregistered - unregistered (-3)
Dec 31 11:36:29 Tower emhttp: Device inventory:
Dec 31 11:36:29 Tower emhttp: shcmd (6): udevadm settle
Dec 31 11:36:29 Tower emhttp: SanDisk_Cruzer_Fit_4C530001201031112561-0:0 (sda) 15297504
Dec 31 11:36:29 Tower emhttp: LOGICAL_VOLUME_50123456789ABCDE_3600508b1001037383941424344450800 (sdb) 976729784
Dec 31 11:36:29 Tower kernel: mdcmd (1): import 0
Dec 31 11:36:29 Tower kernel: mdcmd (2): import 1
Dec 31 11:36:29 Tower kernel: mdcmd (3): import 2
Dec 31 11:36:29 Tower kernel: mdcmd (4): import 3
Dec 31 11:36:29 Tower kernel: mdcmd (5): import 4
Dec 31 11:36:29 Tower kernel: mdcmd (6): import 5
Dec 31 11:36:29 Tower kernel: mdcmd (7): import 6
Dec 31 11:36:29 Tower kernel: mdcmd (8): import 7
Dec 31 11:36:29 Tower kernel: mdcmd (9): import 8
Dec 31 11:36:29 Tower kernel: mdcmd (10): import 9
Dec 31 11:36:29 Tower kernel: mdcmd (11): import 10
Dec 31 11:36:29 Tower kernel: mdcmd (12): import 11
Dec 31 11:36:29 Tower kernel: mdcmd (13): import 12
Dec 31 11:36:29 Tower kernel: mdcmd (14): import 13
Dec 31 11:36:29 Tower kernel: mdcmd (15): import 14
Dec 31 11:36:29 Tower kernel: mdcmd (16): import 15
Dec 31 11:36:29 Tower kernel: mdcmd (17): import 16
Dec 31 11:36:29 Tower kernel: mdcmd (18): import 17
Dec 31 11:36:29 Tower kernel: mdcmd (19): import 18
Dec 31 11:36:29 Tower kernel: mdcmd (20): import 19
Dec 31 11:36:29 Tower kernel: mdcmd (21): import 20
Dec 31 11:36:29 Tower kernel: mdcmd (22): import 21
Dec 31 11:36:29 Tower kernel: mdcmd (23): import 22
Dec 31 11:36:29 Tower kernel: mdcmd (24): import 23
Dec 31 11:36:29 Tower kernel: mdcmd (25): import 24
Dec 31 11:36:29 Tower kernel: mdcmd (26): import 25
Dec 31 11:36:29 Tower kernel: mdcmd (27): import 26
Dec 31 11:36:29 Tower kernel: mdcmd (28): import 27
Dec 31 11:36:29 Tower kernel: mdcmd (29): import 28
Dec 31 11:36:29 Tower kernel: mdcmd (30): import 29
Dec 31 11:36:29 Tower emhttp: import 30 cache device: no device
Dec 31 11:36:29 Tower emhttp: import flash device: sda
Dec 31 11:36:30 Tower emhttp: shcmd (12): /etc/rc.d/rc.avahidaemon start |& logger

 

Again, the above looks good to me.  It sees the sdb connected to my logical drive (which is a raid0).  But then the world falls to fecal matter:

 

Dec 31 11:38:02 Tower emhttp: Device inventory:
Dec 31 11:38:02 Tower emhttp: shcmd (20): udevadm settle
Dec 31 11:38:02 Tower emhttp: SanDisk_Cruzer_Fit_4C530001201031112561-0:0 (sda) 15297504
Dec 31 11:38:02 Tower emhttp: err: ckmbr: read: Input/output error
Dec 31 11:38:02 Tower kernel: sd 1:1:0:0: [sdb] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x01 driverbyte=0x00
Dec 31 11:38:02 Tower kernel: sd 1:1:0:0: [sdb] tag#0 CDB: opcode=0x28 28 00 00 00 00 00 00 00 20 00
Dec 31 11:38:02 Tower kernel: blk_update_request: I/O error, dev sdb, sector 0
Dec 31 11:38:02 Tower kernel: sd 1:1:0:0: [sdb] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x01 driverbyte=0x00
Dec 31 11:38:02 Tower kernel: sd 1:1:0:0: [sdb] tag#0 CDB: opcode=0x28 28 00 00 00 00 00 00 00 08 00
Dec 31 11:38:02 Tower kernel: blk_update_request: I/O error, dev sdb, sector 0
Dec 31 11:38:02 Tower emhttp: device /dev/sdb problem getting size
Dec 31 11:38:02 Tower kernel: Buffer I/O error on dev sdb, logical block 0, async page read

 

And that keeps happening repeatedly.  It likes it the first time, and then each time it tries to do a Device Inventory, it cant seem to read the drive.

 

I also noticed this further up, dont know for sure if it's related:

 

Dec 31 11:36:47 Tower kernel: NMI: PCI system error (SERR) for reason b1 on CPU 0.
Dec 31 11:36:47 Tower kernel: Dazed and confused, but trying to continue

 

That happened between the first good Device inventory and the bad ones following.  Could my P410i be bad?

 

 

 

 

 

 

 

 

 

[1812]

1 hour ago, Carpesimia said:

You mentioned in the other thread that if you intend to use the p410 controller.  Do I have a choice?  With those hot-swap drive bays, dont they all directly connect to the 410 with no alternative?

 

 

I use an h220 hba. You can use any multitude of hba's that unraid can recognize. To use, just disconnect the sas connectors off the motherboard/controller card and reconnect them to that hba, then disable the onboard raid in the bios. works perfectly on the 380's I use to have. It's actually what I had to use on 6.2.4 to maintain easy xfs data retrieval in the event of a drive failure (vs trying to access a raid 0 disk with the data.)

 

 

 

add the community apps plugin (https://forums.lime-technology.com/topic/38582-plug-in-community-applications/ ), then from that, add the unassigned devices plugin and tell me if it appears under unassigned devices. It may have an issue recognizing the unformatted space, and might need to be formatted first via unassigned devices. 

 

1 hour ago, Carpesimia said:

  Could my P410i be bad?

 

I don't think so. They are usually rock solid. To confirm, you're using the P410 onboard, and not a pice version of the p410, correct? in 6.2.4 unRaid HATES hp raid cards that are pcie.

 

Do me a favor and also upgrade to the release client https://forums.lime-technology.com/topic/57820-limetech-release-process-how-do-i-install-prereleases/ . It has solved a lot of issues for me that use to occur on 6.2.4 (and even more in 6.3.x). One of my dl580's runs the onboard p410 raid without issues that plagued it on 6.2.4.

 

 

 

[Carpesimia]

Ok, so I have a little more info.  I was able to start the utility and add all 3 of my drives as individual raid 0 devices.  And yeah, I have the p410i that came with the machine.

 

Anyways, I was able to get unraid to boot, and when i walked into my office, i could actually see all 3 of the drives in the drop-downs.  Hence, I was able to select a parity drive and add other items to the array.  But then I heard a beep from the other room (the server is on the table in the other room) and when i refreshed the screen, the 3 drives were no longer available.

 

I walked into the other room, and saw a kernel error on the screen.  The same error appears in the error log:

 

  1 15:52:26 Tower kernel: NMI: PCI system error (SERR) for reason a1 on CPU 0.
Jan  1 15:52:26 Tower kernel: Dazed and confused, but trying to continue

So the PCI system gives an error, and once that happens the drives are now unreadable.

 

Sounds like something is bad... did the same thing in both versions of unraid that I used.  And it always gave that error, as well, after a minute or two.  Never enough time for the thing to even heat up.

 

 

[1812]

What bios version are you running? (version/date) if its 2015, you may need to try  rolling back or move up to the 6.4RC. I use to get intermittent SERR errors on newer bios on my 380's on 6.2.4 even when I was  using the h220 hba with the p410i enabled in the bios. They stopped when i switched to (i think) 2011/2010 bios and disabled the 410i in bios.

[8064r7]

During boot the p410i probably won't prompt you to press f8 while its booting its own firmware unless you have purposely crashed the system (which I don't recommend).  When you see the prompts for it searching for disks just mash f8 for a while and you will get into its menu where you can delete whatever array the controller had on it at the time and then create all your raid0 stripes for each of your disks.

[Carpesimia]

On 1/1/2018 at 10:06 PM, 1812 said:

What bios version are you running? (version/date) if its 2015, you may need to try  rolling back or move up to the 6.4RC. I use to get intermittent SERR errors on newer bios on my 380's on 6.2.4 even when I was  using the h220 hba with the p410i enabled in the bios. They stopped when i switched to (i think) 2011/2010 bios and disabled the 410i in bios.

 

According to the BIOS, I am running HP BIOS P62 07/02/2013. 

 

I rolled back to the 6.2.4 at your previous request, and get the same errors a few minutes after unraid comes up.  It starts well, but then dies after a few mins.

 

 

[Carpesimia]

11 hours ago, 8064r7 said:

During boot the p410i probably won't prompt you to press f8 while its booting its own firmware unless you have purposely crashed the system (which I don't recommend).  When you see the prompts for it searching for disks just mash f8 for a while and you will get into its menu where you can delete whatever array the controller had on it at the time and then create all your raid0 stripes for each of your disks.

 

The secret to getting into the p410i bios, is during the graphic screen it will at one point say press any key to see messages.  Pressing space will show boot of other devices, including the p410i, and pressing f8 while the raid controller is initializing gets you into the config for it.

[1812]

1 hour ago, Carpesimia said:

 

According to the BIOS, I am running HP BIOS P62 07/02/2013. 

 

I rolled back to the 6.2.4 at your previous request, and get the same errors a few minutes after unraid comes up.  It starts well, but then dies after a few mins.

 

 

 

I sold my 380's before the 6.4RC came out so I never got to try it on those. But I run 6.4RC on a 120G6, z400, and on 2 580G7's and it fixed a few issues (one 580G7 uses P410i onboard exclusively in its setup.) Perhaps try upgrading to the latest release client. Normally I wouldn't recommend it, but it's been solid enough for me to put it on all my hp servers. Proliant support has been growing in unRaid. Not perfect yet, but we're getting there!

 

 

[8064r7]

On 1/3/2018 at 11:54 AM, 1812 said:

 

I sold my 380's before the 6.4RC came out so I never got to try it on those. But I run 6.4RC on a 120G6, z400, and on 2 580G7's and it fixed a few issues (one 580G7 uses P410i onboard exclusively in its setup.) Perhaps try upgrading to the latest release client. Normally I wouldn't recommend it, but it's been solid enough for me to put it on all my hp servers. Proliant support has been growing in unRaid. Not perfect yet, but we're getting there!

 

 

 

Yeah honestly a lot of times within the Linux architecture you are going to run into the same problems, but there has been growing support being added by Limetech and TTPs by the community, so many things are improving for the Proliant enterprise equipment.  6.4 is high enough in RC iterations it does appear to be the best option for anyone wanting to try UnRaid on this equipment.  It will be interesting to see some of the improvements to the Xen implementations in 6.4.

[Carpesimia]

On 1/3/2018 at 12:54 PM, 1812 said:

 

I sold my 380's before the 6.4RC came out so I never got to try it on those. But I run 6.4RC on a 120G6, z400, and on 2 580G7's and it fixed a few issues (one 580G7 uses P410i onboard exclusively in its setup.) Perhaps try upgrading to the latest release client. Normally I wouldn't recommend it, but it's been solid enough for me to put it on all my hp servers. Proliant support has been growing in unRaid. Not perfect yet, but we're getting there!

 

 

 

Ok, so I upgraded to 6.4.0_rc19b.

 

And I still get the error:

 

Jan  4 18:35:22 Tower kernel: NMI: PCI system error (SERR) for reason b1 on CPU 0.
Jan  4 18:35:22 Tower kernel: Dazed and confused, but trying to continue

 

But now, instead of ignoring all 3 of my drives, it only ignores the last one:

Jan  4 18:36:35 Tower kernel: md: unRAID driver 2.9.3 installed
Jan  4 18:36:35 Tower emhttpd: error: get_device_size, 1422: No such device or address (6): open: /dev/sdd
Jan  4 18:36:35 Tower emhttpd: Device inventory:
Jan  4 18:36:35 Tower emhttpd: 3600508b1001037383941424344450800 (sdb) 512 1953459632
Jan  4 18:36:35 Tower emhttpd: 3600508b1001037383941424344450900 (sdc) 512 1953459632
Jan  4 18:36:35 Tower emhttpd: SanDisk_Cruzer_Fit_4C530001201031112561-0:0 (sda) 512 30595072

 

So... progress... but not fixed.  Am I just SOL?  Maybe I should find a different USB bootable image (maybe ubuntu) and see if I still get these errors or if it only happens with unraid.  If it happens in other OS's as well, then its hardware.  Make sense?

 

 

 

[1812]

It makes sense. It may be a bad onboard controller. I forgot about this problem I had : 

 

 

The only other thing to to maybe try is the 2015(?)/last bios for that server. It might help it but will cause an RMRR issue you have to workaround whIch is detailed in the thread in my signature. If you update your bios, be aware that it take a bit of work to roll back unles you have leave the backup bios as original/what you’re on now. 

 

But really, if you put an hba in it and disable the onboard p410i, it’lll be a more “natural” unraid setup with jbod disks. And I believe will eliminate the error. Or at least it did in the one in the thread I posted above.

 

You can test this by disabling the p410i and run a small ssd or 2.5 inch drive off the internal sata port. You may need a specific cable from micro center or frys electronics or some other pc shop/store near you. Might be able to add a second for testing using the cd rom cable, but I don’t remember if that end plug was proprietary or not.

[Carpesimia]

So of I purchase an h220, can I just unplug the drive backplane from the p410i and plug it into the h220 instead?  Is it that simple?

[1812]

50 minutes ago, Carpesimia said:

So of I purchase an h220, can I just unplug the drive backplane from the p410i and plug it into the h220 instead?  Is it that simple?

 

Yup. unplug from the onboard, reroute the sas cables to the pcie card, then during boot, disable the p410i in the bios. I use the h220 in almost all of  mine, but any other hba that works well with unraid will do.

Edited January 6, 2018 by 1812

[Carpesimia]

1 hour ago, 1812 said:

 

Yup. unplug from the onboard, reroute the sas cables to the pcie card, then during boot, disable the p410i in the bios. I use the h220 in almost all of  mine, but any other hba that works well with unraid will do.

 

Thanks so much for being both patient and helpful, it is truly appreciated.  Just made an offer on a H220 on ebay, which I accidentally posted on a different thread.  The good news is someone clued me in that the seller is accepting offers of $30 with free shipping.  So, if he accepts it'll be on its way.

[1812]

6 minutes ago, Carpesimia said:

 

Thanks so much for being both patient and helpful, it is truly appreciated.  Just made an offer on a H220 on ebay, which I accidentally posted on a different thread.  The good news is someone clued me in that the seller is accepting offers of $30 with free shipping.  So, if he accepts it'll be on its way.

 

not a problem. proliants make great unraiding machines (if acquired cheaply!) and I try to help when I can by sharing my mistakes and successes.

 

I use to have 4 380's using unraid as a transcoding cluster for final cut pro x. Was quite fun. Now it's all in a single dl580 with 4 processors. 80 threads of fun! and more pcie slots than you can shake a stick at.

[Carpesimia]

Ok, so here's the latest.

 

Got the H200 on ebay for $30 with free shipping.  Tonight I opened up the proliant and installed the hba into the riser, and connected the backplane.  I booted and in the CMOS disabled the p410.

 

I'm now booted into unraid (only 17 days left  on eval) and can see all 3 drives.  So I selected sdb as my parity, and sdc and sdd as my two usable disks.  I formatted them, but now the parity disk says it's invalid.  But, its also 1% done doing a parity restore (on all brand new drives).  I'm assuming whenever its done restoring nothing, parity will become valid and ill be able to start using unraid.

 

So thats where I am   Will know more probably in the morning.  The restore is going slow as heck, and Im getting tired.. 

[1812]

11 hours ago, Carpesimia said:

(only 17 days left

 

you can usually get an extension or two.

 

11 hours ago, Carpesimia said:

I'm assuming whenever its done restoring nothing, parity will become valid and ill be able to start using unraid.

 

If you don't pre clear disks (zeroing them using the prelcear plugin) then it does make you do a parity check to ensure it can calculate data loss if a problem arrises. But you're on the right path.

 

Also, 6.4 released today, so you can migrate to that. I've already moved all mine over.

[Carpesimia]

So yeah, it finished so everything is in sync.  I now have to add some cache disks.  The machine came with a couple of 140g that should do the trick.  Im not looking for SSD at the moment.

 

How does one upgrade from one version to another?  Ive got the 6.4 pre release now..

[1812]

go to tools>update os, click check for updates on the right side, and then it should show 6.4.0 under the stable branch. 

[Carpesimia]

Sorry for the slow response, but I figured it out about 20 seconds after I asked the question

 

Been running now a couple weeks.  Back today to buy my key.

[1812]

7 hours ago, Carpesimia said:

Sorry for the slow response, but I figured it out about 20 seconds after I asked the question

 

Been running now a couple weeks.  Back today to buy my key.

no worries, have fun!