-
Important Announcement for the TrueNAS Community.
The TrueNAS Community has now been moved. This forum has become READ-ONLY for historical purposes. Please feel free to join us on the new TrueNAS Community Forums
- Status
Daisuke
Contributor
- Joined
- Jun 23, 2011
- Messages
- 1,041
Fair enough. At least we know is coming to a nearby release. Congrats on deciding to use Nginx, next step would be to create a web server plugin. :D
- Joined
- May 28, 2011
- Messages
- 10,996
@Logan54
I think you will have a challenge to run FreeNAS 8 using ZFS with 2GB of RAM. But to answer your original question, yes that was a normal message. Pay close attention to how your system operates, there is a possibility it will crash because you run out of memory. Someone here might be able to tell you how to optimize your system but for now just make sure it doesn't crash.
I think you will have a challenge to run FreeNAS 8 using ZFS with 2GB of RAM. But to answer your original question, yes that was a normal message. Pay close attention to how your system operates, there is a possibility it will crash because you run out of memory. Someone here might be able to tell you how to optimize your system but for now just make sure it doesn't crash.
G
gcooper
Guest
And the question is: how much storage do you have in your box?
Daisuke
Contributor
- Joined
- Jun 23, 2011
- Messages
- 1,041
I guess, more that 2TB. :D
One thing is sure, upgrading from 4GB to 8GB of RAM mixed with a SSD cache on my box gave me a hell of a performance kick (7.8TB on RaidZ2 with 6disks + 1SSD).
Edit: will NFSv4 make it into 8.0.3 stable?
G
gcooper
Guest
Yeah.. having a ZIL helps a lot with sync writes, but RAM is always useful.
No. There's some preliminary support in 8.x, but it's very experimental and I don't recommend using it. If you really need the support, you can turn it on in your KERNCONF (see the nfsv4(4) manpage), but beware -- it might eat your dog -- and the tie-ins in the GUI won't work in some areas.
Daisuke
Contributor
- Joined
- Jun 23, 2011
- Messages
- 1,041
I didn't had the guts to use the SSD for ZIL. I don't want to loose all my terabytes. :)
Plus the controller would not manage well 2 disks on a x1 port (remember our fun discussion about Rocket 620 :)). So I used the SSD as cache instead. I did not get any answer from HighPoint, why am I not surprised. :D
I was thinking to send a card to mav and see if he can test it, but I'm pretty sure he runs older firmware.
G
gcooper
Guest
Oh yeah... duh.. there's that whole ZFS v15 issue where using a non-mirrored ZIL can cause you to almost lose your whole zpool (unless you walk through the complicated brain surgery documented on the FreeBSD forums).
Gcooper,
I haven't seen BETA-3 which I thought you said should be up Thurs/Fri. Just wondering if you were able to fix the broken FUSE package? Taking a peek at the sources I saw some stuff tagged as "RC1", but didn't see anything in the notes about FUSE.
I haven't seen BETA-3 which I thought you said should be up Thurs/Fri. Just wondering if you were able to fix the broken FUSE package? Taking a peek at the sources I saw some stuff tagged as "RC1", but didn't see anything in the notes about FUSE.
G
gcooper
Guest
I'm confused. I haven't posted any images to the SF site yet. I usually switch tags before I wrap things up, then svn copy it over to tags/<version> (I'm not as regimented as jpaetzel in terms of when I do the tagging..).
What Intel NIC do you use (em, igb, ixgb, ixgbe, etc)?
G
gcooper
Guest
That was the original plan, but then I got stuck doing other things. Apparently I need to rebrand it as BETA3 based on Accurus's comments (the changes I applied created some harm unfortunately)...
G
gcooper
Guest
Have an idea of what it might be. What happens if you disable APM/ACPI in the BIOS?
G
gcooper
Guest
Also, what architecture are you running and when does it hang (during the upgrade boot or after that)?
G
gcooper
Guest
1. What does your /boot/loader.conf.local look like?
2. Is it loading em(4) or igb(4) (I can't determine from the e1000 driver file which it would load, but I think it's em(4))?
3. What is your machine model?
4. Does the machine have ipmi support?
I will try to answer as many of these as I can. But some of this is beyond my knowledge as to how to get the information you are looking for. But if you tell me where to look I will get it.
1. On my current working install using BETA 2 there is no data in this file though it does exist.
2. According to the bootup on BETA 2 it is loading: em0
3. By machine model what are you referencing? This is a DIY PC my motherboard model is: Intel DH55TC
4. I have looked up information on my motherboard and it doesn't appear to support ipmi.
1. On my current working install using BETA 2 there is no data in this file though it does exist.
2. According to the bootup on BETA 2 it is loading: em0
3. By machine model what are you referencing? This is a DIY PC my motherboard model is: Intel DH55TC
4. I have looked up information on my motherboard and it doesn't appear to support ipmi.
G
gcooper
Guest
I looked through fresh checkouts of the branches, and the only change that I could possibly see it being is...
What this could do is slow things down by accident, making a race condition that existed before more apparent.
First off, I would try updating your firmware to see whether or not the issue goes away after doing that. If all else fails, let's walk through the following debug decision tree:
When the machine first boots up and you look at the FreeNAS/FreeBSD bootloader (the screen where it says "Welcome to FreeNAS!"), could you hit 6 and...
Procedure 1:
1. Enter in `set hw.em.enable_msix=0'
2. Enter in `boot'.
This will disable MSI-X (it's a more advanced mechanism for receiving PCI interrupts) support for your em chipset.
Procedure 2:
1. Enter in `boot -v'.
This will spew out more messages on the console and might serialize things enough that your boot will succeed. I don't advise sticking with this option because it will fill up /var faster.
Procedure 3:
1. Enter in `set debug.debugger_on_panic=0'
2. Enter in `boot'.
This will disable the kernel debugger at boot.
Procedure 4:
1. Enter in `set hw.pci.enable_msi=0'
2. Enter in `set hw.pci.enable_msix=0'
3. Enter in `boot'.
This will disable all MSI/MSI-X support (they're more advanced mechanism for receiving PCI interrupts). This might cause issues at boot though, depending on the motherboard architecture.
At that point if none of the above methods help or resolve the issue, I would look at issues with interrupt sharing in the BIOS (it's probably not doable because it's a desktop chipset and they tend to dumb that stuff down a bit, but you never know..).
Screenshots for each procedure would help a lot.
1. Same Error as Before
2. Message: MSI IRQ 256 to local APIC 0 Vector 50
Reserved 0x1000 bytes for rid 0x14 type 3 at 0xfe 520000
3. Same Error as Before
4. Still freezes but states: Using Legacy IRQ
5. I tried disabling the NIC to see if it was the only thing holding up the boot. Indeed it appears to just be the NIC. The machine booted fine minus a NIC with the NIC disabled in BIOS.
2. Message: MSI IRQ 256 to local APIC 0 Vector 50
Reserved 0x1000 bytes for rid 0x14 type 3 at 0xfe 520000
3. Same Error as Before
4. Still freezes but states: Using Legacy IRQ
5. I tried disabling the NIC to see if it was the only thing holding up the boot. Indeed it appears to just be the NIC. The machine booted fine minus a NIC with the NIC disabled in BIOS.
- Status
