unRAID OS version 6.4.0 Stable Release Available

[limetech]

4 minutes ago, hmgnsd said:

 

Well, I checked there and the only thing which looks like 'off' to a layman is

 

"Note: After provisioning a Let's Encrypt SSL certificate you may turn off the updatedns and newcert background processes by chaning this field to Yes."

 

Having done that, I'm still getting a 'no vnc' error when I try to use my Windows 10 VM, hence my posting here. Is it off now or what? Has 6.4 broken my VNC?

 

Settings/Identification/SSL Certificate Settings/Use SSL/TLS: set to No, click Apply.  Navigate to any other browser page.

[mifronte]

So I decided to bite the bullet and ended my 234 days of uptime to upgrade successfully to unRAID Pro 6.4.0 from 6.3.5.

 

Since I am running pfSense, I used the pfSense Cert Manager to create an internal certificate for the unRAID server and was able to replace the default certificate with my self-signed certificate and switched the SSL to Yes with no problems.

 

I have a few questions from the log:

 

1. Why is dnsmasq running on my unRAID server?  Do I need it for anything and can it be disabled?

Jan 16 09:58:06 Beanstalk dnsmasq[6602]: started, version 2.78 cachesize 150
Jan 16 09:58:06 Beanstalk dnsmasq[6602]: compile time options: IPv6 GNU-getopt no-DBus i18n IDN2 DHCP DHCPv6 no-Lua TFTP no-conntrack ipset auth no-DNSSEC loop-detect inotify
Jan 16 09:58:06 Beanstalk dnsmasq-dhcp[6602]: DHCP, IP range 192.168.122.2 -- 192.168.122.254, lease time 1h
Jan 16 09:58:06 Beanstalk dnsmasq-dhcp[6602]: DHCP, sockets bound exclusively to interface virbr0
Jan 16 09:58:06 Beanstalk dnsmasq[6602]: reading /etc/resolv.conf
Jan 16 09:58:06 Beanstalk dnsmasq[6602]: using nameserver 10.168.1.1#53
Jan 16 09:58:06 Beanstalk dnsmasq[6602]: read /etc/hosts - 2 addresses
Jan 16 09:58:06 Beanstalk dnsmasq[6602]: read /var/lib/libvirt/dnsmasq/default.addnhosts - 0 addresses
Jan 16 09:58:06 Beanstalk dnsmasq-dhcp[6602]: read /var/lib/libvirt/dnsmasq/default.hostsfile

2. Should I be concerned about this error?

Jan 16 10:17:54 Beanstalk emhttpd: shcmd (235): /usr/bin/php -f /usr/local/emhttp/webGui/include/UpdateDNS.php
Jan 16 10:17:54 Beanstalk root: Error: Nothing to do
Jan 16 10:17:54 Beanstalk emhttpd: shcmd (235): exit status: 1
Jan 16 10:17:54 Beanstalk nginx: 2018/01/16 10:17:54 [alert] 4142#4142: *667 open socket #24 left in connection 11
Jan 16 10:17:54 Beanstalk nginx: 2018/01/16 10:17:54 [alert] 4142#4142: aborting

 

Otherwise, everything appears to be running smoothly.  My UniFi docker is running and my Windows 7 VM is running.  Good job limetech, except for ending my 234 days of uptime!

[hmgnsd]

2 minutes ago, limetech said:

 

Settings/Identification/SSL Certificate Settings/Use SSL/TLS: set to No, click Apply.  Navigate to any other browser page.

 

OK, thank you. That part's done, but I'm still without VNC access to my VM since updating to 6.4.

[hmgnsd]

6 minutes ago, hmgnsd said:

 

OK, thank you. That part's done, but I'm still without VNC access to my VM since updating to 6.4.

 

Oh for god's sake!? It's working but you don't support Safari now!? So I have to Internet Explorer 6 (chrome) to access stuff now!?

Edited January 16, 2018 by hmgnsd

[remati]

On 1/14/2018 at 9:44 AM, limetech said:

 

Try adding to your append line in syslinux:

video=efifb:off

From: https://www.redhat.com/archives/vfio-users/2016-April/msg00236.html

 

Please report if this solves the problem.

 

Adding that to my append line does fix the issue but I managed to get it working also by reverting back to Legacy boot on unraid (opposed to using UEFI).

 

Now I am getting this error on my High Sierra VM using IGD passthrough:

 

Jan 16 09:06:23 Tower kernel: DMAR: DRHD: handling fault status reg 3
Jan 16 09:06:23 Tower kernel: DMAR: [DMA Read] Request device [00:02.0] fault addr 40366000 [fault reason 23] Unknown

 

Any idea what this could be about?

[wgstarks]

11 minutes ago, hmgnsd said:

 

Oh for god's sake!? It's working but you don't support Safari now!? So I have to Internet Explorer 6 (chrome) to access stuff now!?

Actually I think that’s backwards. The bug is in Safari not unRAID. Firefox is also an option.

[deaerator]

Where exactly do I put this information?

 

"Certain motherboards with on-board Aspeed IPMI graphics adaptors may lose video or switch to different colored text during the boot process. To prevent this, add the nomodeset kernel option to your syslinux append line."

[pwm]

10 hours ago, HellDiverUK said:

So many questions.  It's almost like people never bother looking at any of the threads in the pre-release forum.

People should not have to look at threads in the pre-release forum. That's for beta-testers.

 

Common questions from that sub-forum should either have become FAQ entries in this thread - or not be needed anymore because the stable release should contain changes based on the problems people had during the pre-release stage.

 

But it's more bothersome that people ignore the content of the first post in this very thread - lots of people who jump in and wonder why their custom http port doesn't work anymore despite being mentioned in the release notes. Besides showing a serious lack of focus, it also shows that people don't care/worry very much about their data and system functionality - they don't check if the update might introduce any problematic change that is incompatible with their current use of the system.

[trurl]

15 minutes ago, deaerator said:

Where exactly do I put this information?

 

"Certain motherboards with on-board Aspeed IPMI graphics adaptors may lose video or switch to different colored text during the boot process. To prevent this, add the nomodeset kernel option to your syslinux append line."

 

I searched this very thread and found this for one example:

 

[pwm]

4 hours ago, Fireball3 said:

You can still use the old way without SSL.

 

But I'm also wondering, what's the intended use case of the whole "secure access" thing, if the

server is in a trusted environment (LAN)?

 

Insert one (1) Windows machine in your trusted environment.

 

Give that Windows machine network access.

 

You now no longer have a trusted environment.

 

Good practice in a network is to configure every single machine as if even your own, local, machines are hostile. It saves lots of troubles at a later stage if your trusted machine for some reason happens to become hostile - intentionally or through some bug.

 

That's the same type of thinking that is the reason why people should not run as Administrator or root for their everyday work. The access rights you haven't given yourself will not allow you to accidentally break things.

[deaerator]

New Problem

 

During boot up, the array doesn't start up.

I managed to SSH and retrieve the logs.

tower-diagnostics-20180116-1527.zip

[1812]

46 minutes ago, pwm said:

\

But it's more bothersome that people ignore the content of the first post in this very thread - lots of people who jump in and wonder why their custom http port doesn't work anymore despite being mentioned in the release notes. Besides showing a serious lack of focus, it also shows that people don't care/worry very much about their data and system functionality - they don't check if the update might introduce any problematic change that is incompatible with their current use of the system.

 

that's because everyone is use to updating their iPhones and not reading anything about the update. and they then assume that anything else that wants to update would be the same essentially.

[pwm]

1 minute ago, 1812 said:

 

that's because everyone is use to updating their iPhones and not reading anything about the update. and they then assume that anything else that wants to update would be the same essentially.

And people regularly gets burned when updating phone apps too - they auto-update and then find that some features have been removed and are only available in a new premium version.

 

I think we would be able to find a rather high correlation between people who ignore reading release notes and people who install new equipment and leave them with the factory-default passwords.

[scottc]

Upgraded without issue

[Shinobi]

On 1/15/2018 at 7:23 AM, limetech said:

There are issues with Ryzen, try disabling "Global C-state control" in your bios.

 

Are they not already fixed with unRaid 6.4rc7 ?

as explained in this post ?

 

On 2/22/2017 at 9:26 PM, lionceau said:

Now that Ryzen is official I'm quite interested in it.

 

Upto 8 Cores/16 Threads, excellent SMT implementation, ECC RAM support, IOMMU, 8 port on-CPU SATA with the X370 chipset and the high end models are half the price of comparable Intel CPUs.

 

It seems to tick all boxes but I'm concerned whether it'll play nice with unRAID and VM PCI/GPU passthrough. Is anybody planning to test this?

 

 

Update to make things a bit easier for people interested in running Unraid on Ryzen:

 

unRAID 6.4rc7+ generally works fine with Ryzen as of July 2017. Be sure to follow your motherboard manufacturers instructions to update to the latest UEFI version with AGESA 1.0.0.6.

 

As of right now (6.4rc6) there appears to be one bug related to CPU power-managment that causes unRAID to crash intermittently. 

Good news: This bug was fixed in 6.4rc7! 

The way to fix this is to either disable "Global C-State Control" in the motherboard's UEFI/BIOS or have a constant load on the machine (i.e. with a Windows VM idling) so that the CPU never throttles frequency down completely at idle. Either of these fixes will lead to slightly increased power consumption.

 

The second bug is in KVM and leads to decreased video performance when passing through GPUs into VMs on Ryzen. The problem is related to Nested Page Tabling; Ryzen owners have observed decreased CPU and I/O performance inside the VM when NPT is off and decreased FPS in graphic intensive applications (i.e. video games) when NPT is on.

Gaming performance inside a unRAID VM with passed through GPU can be disappointing for that reason. Some systems seem to be affected more than others. The good news is that Xen hypervisor is not affected by this problem so it looks like the bug is in the KVM software, not AMD's hardware;  KVM lead developers have acknowledged the issue but are lacking time/resources and did not mention a timeframe for a fix.

There are some more advanced workarounds that can improve the situation by turning off NPT and passing through entire storage controllers but don't work around it completely.

 

If you are affected by this issue please take a look at https://bugzilla.kernel.org/show_bug.cgi?id=196409 and consider commenting to raise awareness among developers and perhaps even AMD.

 

 

 

[JustinChase]

I upgraded today from rc20 (I think), and when I rebooted my server, it never came back up.  I can't connect to it via the IP, nor with Putty.  I hard rebooted it (a couple times), and still cannot reach the server.

 

I then connected my monitor, and I can't see that it's even trying to do anything.  None of the HDMI ports display anything on my monitor, they all result in the monitor saying no input found.

 

I've removed the USB, and downloaded directly from limetech servers, checked the MD5 is correct and ran a check disk on the USB.  I've put it back in the server, and booted, and I still get nothing on the monitor.  not with the motherboard HDMI (which is what used to work), nor from either of the 2 video cards.

 

I really don't know what the next troubleshooting step is here.

[JorgeB]

I really don't know what the next troubleshooting step is here.

Do you see or can get into the motherboard bios? If not the problem isn't unRAID, if yes when do you lose video output?

 

 

[djg85]

And here was me being parranoid and removing some of the "known" problematic plugins mentioned in this thread prior to rebooting

 

Anyone else using ESXi have issues updating? 

 

 

I get the same error whether I try and boot into safemode or not.  Everything was running fine before the reboot/upgrade.  I'm on ESXi 6.5.0 and PLOP.

[Living Legend]

I just wanted to note here that I tried probably 75% of the pre releases, and then the stable 6.4 release.  All failed with a hard crash within an hour.

 

After running in safe mode and narrowing it to plugins, I have now been up and running for 2 days after removing the Dynamix Sleep Plugin.

 

Web UI is super smooth and quick.  Only thing I don't like is that there is no padding on the edge and it looks like something you'd have seen on the web in the late 90s. 

[FreeMan]

I upgraded on Sunday (the 14th), and everything went swimmingly. I woke up Monday morning (the 15th) to find a Pushover notification at about 04:30 that a parity check started followed a few minutes later by one indicating that FCP had detected an unclean shut down. That's pretty unusual for me as I usually have >100 days uptime and only bring the server down for upgrades (software or hardware). I let it ride since I had to head to work.

 

I woke up this morning (the 16th) to find that the server had restarted at about 01:30 and launched another parity check. I had enough presence of mind to put FCP into trouble shooting mode, which was handy because at about 08:30 this morning I got another set of notices that a parity check had started and FCP had detected an unclean shut down.

 

Before I left this morning, I shut down all dockers except OpenVPN so I'd be able to check in on it from work during the day. I was receiving tons of errors indicating issues with the Advanced Buttons plugin, so I connected via TeamViewer and uninstalled that.

 

Logs are attached. Server specs in my .sig are current. I haven't had a chance to run a memtest yet - I'm hoping to make it through a parity check before taking the system down for that. The parity check has been running almost 11 hours now and is about 50% complete (yes, it's normally that slow - that's another issue to tackle).

 

I'm open to any and all suggestions.

FCPsyslog_tail.txt

nas-diagnostics-20180116-0816.zip

[h3xcmd]

 

Failed to Allocate memory for Kernel command line, bailing out
booting kernel failed: bad file number

-UnRaid Trial- 18days remaining

1. Created a 6.3.5 USB using the USB creator tool for windows. Started the Trial for 2 weeks with MBR BIOS boot. 
2. Saw that the new stable 6.4.0 was out, used the WebGUI to update
3. Reboot required
4. Upon loaded WebGUI enabled UEFI in the Flash media on main page (so far so good) Reboot
5. (Set Server BIOS settings to default) Changed my Servers BIOS system setting to boot to UEFI
6. Deleted previous UEFI boot menu paths (Windows, ESXi..etc) 
7. Created a UEFI boot entry pointing to /EFI/Boot/Boot64.efi (1st attempt I did not create a UEFI boot path same result)
8. Rebooted server
9. UnRaid Boots up in UEFI
10. Auto boot count down
11.      loading bzroot...ok
12. "Failed to Allocate memory for Kernel command line, bailing out
    booting kernel failed: bad file number"
     

I had this same issue prior to starting the trial creating a 6.4.0RC19 Boot usb. Just I knew there was not going to be any support for a BETA now that 6.4.0 is stable any Ideas or possible solutions. 

I should also note that all of the boot options including memtest all have the same failed message as above. 

I can only Tab to edit >/bzimage initrd=/bzroot_

I really don't want to be stuck using legacy boot mode. Thanks in advance to any suggestions or help. 

[EDIT] 01/16/2018 @9:55p
So after a little research apparently This is an old issue from 05/2015 with Syslinux and Dell Poweredge R410. T420, R320 servers, Though why would esxi Work, it is also syslinux? I also Started a New thread in General Support on this topic, If you would like to help or provide any leads please head over there. Thansk!

> the message comes from main.c when allocate_pages() fails.
> allocate_pages() is just a wrapper that calls BS->AllocatePages.

http://www.syslinux.org/archives/2015-April/023398.html
http://www.syslinux.org/archives/2015-May/023466.html
https://github.com/geneC/syslinux/blob/master/efi/main.c

I

 
-----------------------------
Dell PowerEdge T420  | 24gb ECC Ram | 2x 12core Xeon E5-2420 | 5 Drives (1TB parity | 3TB | 500GB cache)

 

Edited January 17, 2018 by h3xcmd

[1812]

1 hour ago, Living Legend said:

Web UI is super smooth and quick.  Only thing I don't like is that there is no padding on the edge and it looks like something you'd have seen on the web in the late 90s. 

 

you do know there is more than one web gui skin, right?

[1812]

also, updated a t3400 without issue this evening. but it's basically a no frills backup nas. It can't identify the processor, which is interesting. it's a q9650, and the log shows

 

Jan 16 16:23:29 Backup kernel: smpboot: CPU0: Intel(R) Core(TM)2 Quad CPU    Q9650  @ 3.00GHz (family: 0x6, model: 0x17, stepping: 0xa)
Jan 16 16:23:29 Backup kernel: Performance Events: PEBS fmt0+, Core2 events, Intel PMU driver.

 

 

 

but the information tab shows as CPU: Not Specified @ 3000. it doesn't appear to be anything tragic i suppose.

Edited January 17, 2018 by 1812

[cpthook]

22 hours ago, cpthook said:

Updated both my servers today, however, one of them will not allow access to the GUI.  I've tried each boot option but no success.  I've included diagnostics and any assistance would be welcomed.  I've reverted back to v6.3.5 so I'm back to normal for now, until I can figure out why I cant access the web GUI with 6.4.    

 

towerii-diagnostics-20180115-1912.zip

Bumping my post hoping for some assistance.  I did not have custom http settings with v6.35 except for a static IP, so I'm puzzled as to why no Web GUI after updating to v6.4.  Aafter update, my machine is pinging on my LAN both with IP and hostname.  My go script is configured per the recommendations from the OP and network config files are the same as before.  Thanks for any assistance provided

[h3xcmd]

1 hour ago, Living Legend said:

Web UI is super smooth and quick.  Only thing I don't like is that there is no padding on the edge and it looks like something you'd have seen on the web in the late 90s. 

Azure Dynamix color theme is pretty nice to look at.