New user, new install: FreeNas boot errors (message file)

Status
Not open for further replies.

Horia Georgescu

Dabbler
Joined
Aug 23, 2014
Messages
11
Hello;
I tried a while back to use FreeNas for a home NAS, and didn't get to far. I took a break and now came back to it. My hardware AMD based motherboard and 8 core CPU (M5A99X-EVO-R20-ASUS), 16 GB ECC RAM, an LSI Megaraid controller flashed with IT firmware. I did a test install on one SSD (I have two SSDs in the system). I only created the root pool (zfs on root), saving for now the data hard drives (my hard drives are disconnected until I have a stable boot configuration, then I'll create a couple of pools for my media).

I'm attaching my messages file and I've also pulled from it a few areas of concern. I'm not sure if anyone has some thoughts on how to deal with them (aside from getting new hardware ), so that I can use FreeNas.

I know that my hardware works properly, having loaded current versions of linux and freebsd/pc-bsd onit and I had no issues.


SATA driver errors:

Feb 7 03:58:06 freenas (aprobe0:ahcich2:0:0:0): CAM status: ATA Status Error
Feb 7 03:58:06 freenas (aprobe0:ahcich2:0:0:0): CAM status: ATA Status Error
Feb 7 03:58:06 freenas (aprobe0:ahcich3:0:0:0): CAM status: ATA Status Error
Feb 7 03:58:06 freenas (aprobe0:ahcich3:0:0:0): CAM status: ATA Status Error

LSI driver a few versions behind the firmware:

Feb 7 03:58:06 freenas mps0: <LSI SAS2008> port 0xd000-0xd0ff mem 0xfe5c0000-0xfe5c3fff,0xfe580000-0xfe5bffff irq 28 at device 0.0 on pci2
Feb 7 03:58:06 freenas mps0: Firmware: 20.00.00.00, Driver: 16.00.00.00-fbsd
Feb 7 03:58:06 freenas mps0: IOCCapabilities: 1285c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,EventReplay,HostDisc>

CPU power management errors:
Feb 7 03:58:06 freenas acpi_throttle1: failed to attach P_CNT
Feb 7 03:58:06 freenas acpi_throttle2: failed to attach P_CNT
Feb 7 03:58:06 freenas acpi_throttle3: failed to attach P_CNT
Feb 7 03:58:06 freenas acpi_throttle4: failed to attach P_CNT
Feb 7 03:58:06 freenas acpi_throttle5: failed to attach P_CNT
Feb 7 03:58:06 freenas acpi_throttle6: failed to attach P_CNT
Feb 7 03:58:06 freenas acpi_throttle7: failed to attach P_CNT

These can be dealt with by adding these directives to loader.conf:
hint.acpi_throttle.0.disabled=1
P4TCC with hint.p4tcc.0.disabled=1

Other errors/warnings:

Feb 7 03:58:06 freenas WARNING: VIMAGE (virtualized network stack) is a highly experimental feature.
Feb 7 03:58:06 freenas ACPI Warning: FADT (revision 5) is longer than ACPI 5.0 version, truncating length 268 to 256 (20111123/tbfadt-325)
Feb 7 03:58:06 freenas ACPI Warning: Optional field Pm2ControlBlock has zero address or length: 0x0000000000000000/0x1 (20111123/tbfadt-606)
Feb 7 03:58:06 freenas root: /etc/rc: WARNING: failed to start watchdogd
Feb 7 03:58:07 freenas root: /etc/rc: WARNING: Dump device does not exist. Savecore not run.
Feb 7 03:58:09 freenas root: /etc/rc: WARNING: failed precmd routine for vmware_guestd
Feb 7 03:58:17 freenas generate_smb4_conf.py: [common.pipesubr:58] Popen()ing: /usr/local/bin/net sam rights grant WARNING SeTakeOwnershipPrivilege SeBackupPrivilege SeRestorePrivilege
 

Attachments

  • messages.txt
    22.9 KB · Views: 308
Last edited:

BigDave

FreeNAS Enthusiast
Joined
Oct 6, 2013
Messages
2,479
LSI driver a few versions behind the firmware: Feb 7 03:58:06 freenas mps0: Firmware: 20.00.00.00, Driver: 16.00.00.00-fbsd
Actually your firmware is a few versions AHEAD of your driver provided by Freenas.
Downgrade controller firmware to version P16 if you expect proper behavior in Freenas.
Here's a link to the firmware area at LSI website. About half way down, expand Firmware,
click on the ARCHIVE link to find the P16 downloads.
 

Horia Georgescu

Dabbler
Joined
Aug 23, 2014
Messages
11
Thanks, BigDave. You can look at it the other way too :) the LSI driver (in FreeNAS ) is a few versions behind the firmware in MegaRaid card.
I guess there is no easy way to upgrade the freenas (freebsd) driver with the current version (20) on LSI site?
 

Ericloewe

Server Wrangler
Moderator
Joined
Feb 15, 2014
Messages
20,194
Thanks, BigDave. You can look at it the other way too :) the LSI driver (in FreeNAS ) is a few versions behind the firmware in MegaRaid card.
I guess there is no easy way to upgrade the freenas (freebsd) driver with the current version (20) on LSI site?

That's not an option. If it were plug and play, you'd still have to rebuild FreeNAS from source. And it's not just a matter of replacing the old driver with the new one.
 

BigDave

FreeNAS Enthusiast
Joined
Oct 6, 2013
Messages
2,479
I guess there is no easy way to upgrade the freenas (freebsd) driver with the current version (20) on LSI site?
As Eric has pointed out, it's not an option, please accept this and take your firmware BACK to version P16.
The reality is that P16 driver is just as performance tuned and feature packed as the newer ones (it's true),
and works best with FreeBSD/FreeNAS.
 

Horia Georgescu

Dabbler
Joined
Aug 23, 2014
Messages
11
Ok then.
Will do that. Thank you.

Do you have any thoughts about the other messages I'm getting in the log?
 

BigDave

FreeNAS Enthusiast
Joined
Oct 6, 2013
Messages
2,479
Your attached file is not readable so it's not going to be of any value.
Fix your controller's firmware and then after it's applied,
post another log message with proper formatting.
 

Horia Georgescu

Dabbler
Joined
Aug 23, 2014
Messages
11
Sorry - just tested and the log file is ok. I had to add the txt extension (otherwise would have been just "messages"), to be able to upload it into the forum. Saved locally on the hard drive opens just fine with less/vi, or any of the GUI editors.
Still have to get the firmware issue resolved. Will come back with updates.
If there are any suggestions about the other messages (which are not related to the mps driver), would appreciate.
Thank you.
hg
 

Horia Georgescu

Dabbler
Joined
Aug 23, 2014
Messages
11
Hello again.
After a bit of struggle, been able to downgrade the card to P16.
Disconnected all the other drives, installed freenas on a 32GB USB stick and booted.

Lots of strange messages at boot-up.
Have initiated a few updates from the console and the web interface.
Still lots of things the system doesn't like, after these updates. The usb install and boot-up is slow. very slow.
I learned in the process that the USB 3.0 is disabled, for some considerations of past unreliable experiences. I would have liked to try to go through the install with and without USB3.0, but I don't know how to do that. After the system is installed, I can go modify the loader.conf file, but I didn't yet.

I thought it useful to record the entire boot process, since dmesg file is not ctaching the very early sequence.
I held my phone in front of the screen, and recorded one such boot cycle.
My hand wasn't too steady, but is not too bad. Can be stopped and looked those messages. Here is the link to the mov file on dropbox: https://www.dropbox.com/s/wjoxn5xk5abxbmy/IMG_1331.MOV?dl=0.
I'll attach the dmesg file shortly, or share it on dropbox too.

The wizard hangs very early on, and had to exit it. Ideally would have to have a way to bypass certain sections, if they don't make sense at the time of the boot. I suppose that the absence of hard drives is confusing, but for the beginning I would still want to be able to setup the timezone, the domain, and a few other things from the web interface. I want to install hard drives only after I get rid of those pesky messages.
All in all, not a stellar experience. I remembered why I gave up in summer last year.

The interface, the features and the entire package is so outstanding, it is just such a pity that it doesn't seem to work reliably for me/on my hardware.

Still hoping there are some other things I can try. Any other suggestions from the seasoned FreeNas guru-s on this forum?

Thank you.
HG
 
Joined
Mar 6, 2014
Messages
686
I have looked at your recording. I am by no means a guru, but it seems to boot just fine.
 

danb35

Hall of Famer
Joined
Aug 16, 2011
Messages
15,504
Hello again.
The interface, the features and the entire package is so outstanding, it is just such a pity that it doesn't seem to work reliably for me/on my hardware.
AMD doesn't have very good support for FreeBSD, or you could say instead that FreeBSD doesn't have very good support for AMD. That fact is pretty well-known and well-documented. That may or may not have anything to do with the issues you're seeing, though. What specific "pesky messages" are you concerned about?
 

Glorious1

Guru
Joined
Nov 23, 2014
Messages
1,211
Minor thing, but I think you could shorten the boot time a bit by erasing the BIOS on the LSI controller.
 

Horia Georgescu

Dabbler
Joined
Aug 23, 2014
Messages
11
Thanks Everyone for chiming in.
@danb35: Thank you. I only discovered this reality after I've purchased all the hardware. hindsight is 20:20. But this was originally meant to run ESXi for me (I found references that it was capable of that, and that indeed worked). But ESXi added a level of complexity I think I don't need/nor have the time to manage.
So I stayed /tried to stay on the low cost side ... and I don't know I would have jumped on a Intel CPU & Motherboard, knowing that there was reduced support from FreeBSD.
@Glorious1: I will remove the LSI bios from the controller - did it once and it was too quiet (I mean I wasn't sure it was working properly), and installed it back. But now the bios it is extending so much the boot time.
@Rilo Ravestein: There is a lot of stuff in the boot messages, and it is true not a lot is bad. There are just a handful of things which the boot process really complains about, and I'm not sure of their importance.

Today I've just reinstalled freenas again, and recorded the second boot after the installation.
I've also picked up an old HDD, and have it plugged in the system, on the on-board SATA, but for the moment I have not attached it to freenas. (there are a couple of smartd errors relative to the HDD - I know those are due to the age of the drive. In the messages file they show up at the end.

But for now all my focus is what is happening with FreeNas OS on the USB memory stick.

I booted again - recorded - and prompted a system update from the console. recorded all the way, including the reboot. Saved everything on Dropbox. The names of the mov-files show the sequence they came in.
https://www.dropbox.com/sh/oabbxkybln5voc4/AADsCSUid4YGCHv7RzhzqDzGa?dl=0

Added the log in the same directory (/var/log/messages file) as well, on dropbox. Messages recorded during last reboot.

Here I'm pulling out from the message file the warnings and/or errors of any sort and include below, for discussion. I've made a couple of comments prefaced by [HG] where I know something about the respective line.

Feb 13 09:06:04 freenas WARNING: VIMAGE (virtualized network stack) is a highly experimental feature.
Feb 13 09:06:04 freenas ACPI Warning: FADT (revision 5) is longer than ACPI 5.0 version, truncating length 268 to 256 (20111123/tbfadt-325) [HG]: apparently this is harmless?
Feb 13 09:06:04 freenas ACPI Warning: Optional field Pm2ControlBlock has zero address or length: 0x0000000000000000/0x1 (20111123/tbfadt-606)


Feb 13 09:06:04 freenas kbd0 at atkbd0
Feb 13 09:06:04 freenas atkbd0: [GIANT-LOCKED]
Feb 13 09:06:04 freenas wbwd0: HEFRAS and EFER do not align: EFER 0x2e DevID 0xff DevRev 0xff CR26 0xff

Feb 13 09:06:04 freenas acpi_throttle0: <ACPI CPU Throttling> on cpu0 [HG]: these are correctable with a couple of flags in loader.conf, disabling the power mgmt on the CPU. Apparently harmless?
Feb 13 09:06:04 freenas hwpstate0: <Cool`n'Quiet 2.0> on cpu0
Feb 13 09:06:04 freenas acpi_throttle1: <ACPI CPU Throttling> on cpu1
Feb 13 09:06:04 freenas acpi_throttle1: failed to attach P_CNT
Feb 13 09:06:04 freenas device_attach: acpi_throttle1 attach returned 6
Feb 13 09:06:04 freenas acpi_throttle2: <ACPI CPU Throttling> on cpu2
Feb 13 09:06:04 freenas acpi_throttle2: failed to attach P_CNT
Feb 13 09:06:04 freenas device_attach: acpi_throttle2 attach returned 6
Feb 13 09:06:04 freenas acpi_throttle3: <ACPI CPU Throttling> on cpu3
Feb 13 09:06:04 freenas acpi_throttle3: failed to attach P_CNT
Feb 13 09:06:04 freenas device_attach: acpi_throttle3 attach returned 6
Feb 13 09:06:04 freenas acpi_throttle4: <ACPI CPU Throttling> on cpu4
Feb 13 09:06:04 freenas acpi_throttle4: failed to attach P_CNT
Feb 13 09:06:04 freenas device_attach: acpi_throttle4 attach returned 6
Feb 13 09:06:04 freenas acpi_throttle5: <ACPI CPU Throttling> on cpu5
Feb 13 09:06:04 freenas acpi_throttle5: failed to attach P_CNT
Feb 13 09:06:04 freenas device_attach: acpi_throttle5 attach returned 6
Feb 13 09:06:04 freenas acpi_throttle6: <ACPI CPU Throttling> on cpu6
Feb 13 09:06:04 freenas acpi_throttle6: failed to attach P_CNT
Feb 13 09:06:04 freenas device_attach: acpi_throttle6 attach returned 6
Feb 13 09:06:04 freenas acpi_throttle7: <ACPI CPU Throttling> on cpu7
Feb 13 09:06:04 freenas acpi_throttle7: failed to attach P_CNT
Feb 13 09:06:04 freenas device_attach: acpi_throttle7 attach returned 6

Feb 13 09:06:04 freenas Trying to mount root from zfs:freenas-boot/ROOT/FreeNAS-9.3-STABLE-201502110455 []...
Feb 13 09:06:04 freenas GEOM_RAID5: Module loaded, version 1.3.20140711.62 (rev f91e28e40bf7)
Feb 13 09:06:04 freenas wbwd0: HEFRAS and EFER do not align: EFER 0x2e DevID 0xff DevRev 0xff CR26 0xff
Feb 13 09:06:04 freenas root: /etc/rc: WARNING: failed to start watchdogd
Feb 13 09:06:06 freenas root: /etc/rc: WARNING: Dump device does not exist. Savecore not run.
Feb 13 09:06:07 freenas root: /etc/rc: WARNING: failed precmd routine for vmware_guestd

That's about it in the messages file...

Aside from this, I've already mentioned before the issue with the wizard - but will see if it still hangs, if I have now a HDD attached to the system.

@BigDave: the fact that I had to go back to Phase 16 on the controller is still bugging me (I'm thinking that there must have been necessary changes, between P16 and P20). I understand that the community experience with P16 was good, therefore I will have to reconcile for myself this area.
 
Last edited:

Horia Georgescu

Dabbler
Joined
Aug 23, 2014
Messages
11
I just realized that those CAM errors I reported in the beginning, are gone. The difference is that I disconnected the SSD drives, and connected an old SATA to it. Is it that the CAM driver is not "talking" properly to the on-board controller, talking to the SSD drives? This must be the explanation. Both the MB and The SSDs are in perfect state....
 
Status
Not open for further replies.
Top