February 6, 20188 yr On 2/4/2018 at 8:08 PM, CHBMB said: It's pretty valid "hysteria" as far as I can tell... Admittedly a home server running unRAID may not be the most obvious of targets, but I think the concern has been pretty justified. Don't know how as it is pretty much back-end without user download activity.. and Linux Anyway, installed OK, took few minutes.
February 8, 20188 yr FYI - This was an upgrade from v6.3.5 direct to v6.4.1. Someone pointed me to this thread (really wish I'd seen it before upgrading!) and the Aspeed IMPI fix got video working again for my ASRock motherboard. It would boot up but at the point the login prompt should appear the screen would just go blank, which was irritating but at least the web GUI and SSH still worked. The fix just has you add the "nomodeset" parameter to the main/default unRAID OS label/boot option's append line. Do I need to or should I add that "nomodeset" parameter to the other labels as well - GUI and Safe Modes? Just want to be sure I won't run into the problem again, especially if I need to go in to safe mode. Another issue I have is that after the upgrade I can't get a VM to boot. My VM Manager starts up fine, so it isn't to do with the common problem of the paths getting lost or messed up mentioned in the OP. But when I try to start my Windows 10 VM after the upgrade I get bluescreen/frowny face with "your pc ran into a problem and needs to restart" in the VM. It proceeds to do so but ends u p just going to a "your pc did not start correctly" screen. Shutting down and rebooting it the same thing happens again and again. I tried a startup repair from that screen, but that didn't work. The only other thing I've tried changing so far is upping the q35-2.7 to the latest q35-2.10. No luck there either. Attaching the [removed xml] for that VM. Frustrating as that VM was chugging along fine just a couple hours ago under v6.3.5. My Windows 7 VM seems to be doing just fine, but it is running seabios+i440fx rather than the Win10 one above which uses ovmf+q35, though I'm not sure that makes a difference or not. Also have a CentOS 7 VM that comes up fine and that uses seabios/q35. None have any physical hardware passed through to them. I've made a quick copy of my disk *.img files for the VMs in case anything corrupts my Windows install hopefully I can revert back if needed. edit: Got it to boot by deleting the VM and recreating it, pointing back to my same .img file again for the C: drive. I remembered I made some tweaks to that VM from this thread a while back to do with SSD optimization and best guess is those settings messed things up with the new version of unRAID/KVM. Leaving this here though in case someone runs into the same issue. Edited February 8, 20188 yr by deusxanime hopefully got it fixed on my own
February 18, 20188 yr Hello, why im getting this message? ive checked this topic and its mention to add zenstates do disable - from which reason? my system is stable, thanks
February 18, 20188 yr You're getting the message because the upgrade notes recommend to add that line to all Ryzen systems. Whether or not you need to do so or not, I cannot answer.
February 18, 20188 yr 58 minutes ago, Squid said: You're getting the message because the upgrade notes recommend to add that line to all Ryzen systems. Whether or not you need to do so or not, I cannot answer. @killeriq, if you don't add it, write yourself a BIG note and post on your server that you didn't do it. That way if you do get a problem in the future, you will have a reminder of the first thing to try!
February 18, 20188 yr 1. Is there need to install something or the "line in config" is that "installation" ? 2. does is das to do anything with C-states enable/disable in BIOS ?
February 18, 20188 yr 26 minutes ago, killeriq said: 1. Is there need to install something or the "line in config" is that "installation" ? 2. does is das to do anything with C-states enable/disable in BIOS ? Your go file in the config directory/folder needs to look like this. Be sure that you use a Linux aware editor (not Notepad that comes with Windows!). #!/bin/bash # Start the Management Utility /usr/local/sbin/zenstates --c6-disable /usr/local/sbin/emhttp & For number 2, This setting is somewhere in your BIOS which you access before the unRAID boot process starts. You may have to read the MB manual to find out the proper key to hit to access the BIOS.
February 18, 20188 yr 1 minute ago, Frank1940 said: Your go file in the config directory/folder needs to look like this. Be sure that you use a Linux aware editor (not Notepad that comes with Windows!). #!/bin/bash # Start the Management Utility /usr/local/sbin/zenstates --c6-disable /usr/local/sbin/emhttp & For number 2, This setting is somewhere in your BIOS which you access before the unRAID boot process starts. You may have to read the MB manual to find out the proper key to hit to access the BIOS. 1. thanks 2. i have them ENABLED in BIOS, know where is it , thats why i wonder why to disable "/usr/local/sbin/zenstates --c6-disable" if is related or not...I know they were issues when ENABLED in below v6.40RC , but now is fine, no system freeze
February 20, 20188 yr Is there still an issue with the temperature readings from NVMe drives not being detected?
February 21, 20188 yr Author Hi folks, please use this thread to discuss the update notes (i.e. the OP) themselves. If you need support or have questions about the update, the actual release thread would be a better place to post.
March 20, 20188 yr I would suggest adding information about regarding this situation: https://lime-technology.com/forums/topic/70095-unraid-os-version-650-stable-release-available/?page=8&tab=comments#comment-644097 It seems like more and more MB's are having problems booting using legacy mode. Pointing it out could save a lot of threads and posts about the issue...
March 20, 20188 yr 3 hours ago, Frank1940 said: It seems like more and more MB's are having problems booting using legacy mode. Pointing it out could save a lot of threads and posts about the issue... Interestingly, with all prior versions of unRAID, booting in legacy mode with this board (ASRock C236 WSI) was no problem. Something changed in unRAID 6.5.0 that no longer allows legacy boot. Limetech is not sure what changed as they did nothing that, to their knowledge, would impact that. However, since a roll-back to 6.4.1 once again boots the board in legacy mode, something must have changed in the Linux kernel with the current release. Perhaps other boards will be similarly affected in the future. The fix for this board involves making sure UEFI boot is the ONLY boot option in the BIOS. If other boot options exist, it fails to boot even if UEFI is the first priority. Edited March 20, 20188 yr by Hoopster
March 21, 20188 yr Author thanks @Frank1940 and @Hoopster, I added details for: getting stuck "loading /bzroot" NETBIOS names, which characters are valid How to access VM XML View
March 21, 20188 yr On 3/20/2018 at 8:34 AM, Hoopster said: The fix for this board involves making sure UEFI boot is the ONLY boot option in the BIOS. If other boot options exist, it fails to boot even if UEFI is the first priority. That's interesting because for my board pure EFI or pure Legacy booting causes problems with my virtual machines, but setting the bios to, "EFI mode with Legacy support" (CSM on some boards) and then making sure unRAID boots as a legacy OS, is what makes my setup happy. Then again I've been on this mobo (new to me board) since only 6.4.0-rc's. I would agree there kernel/(a feature) is hunting for something out of the bios, and boot-mode seems to effect whether it gets the resource, or not.
March 21, 20188 yr 5 minutes ago, Jcloud said: That's interesting because for my board pure EFI or pure Legacy booting causes problems with my virtual machines, but setting the bios to, "EFI mode with Legacy support" (CSM on some boards) and then making sure unRAID boots as a legacy OS, is what makes my setup happy. Then again I've been on this mobo (new to me board) since only 6.4.0-rc's. I would agree there kernel/(a feature) is hunting for something out of the bios, and boot-mode seems to effect whether it gets the resource, or not. Just to clarify, I am referring to UEFI: (flash dirve) boot being the only option in the boot priority settings. CSM is enabled and the boot option filter is set to "UEFI and Legacy." My full findings are here:
March 24, 20188 yr Hi, I've run the upgrade assistant to upgrade from 6.41 to 6.5 and it gives me this message: Issue Found: Cache drive partition doesn't start on sector 64. You will have problems. See here https://lime-technology.com/forums/topic/46802-faq-for-unraid-v6/?tab=comments#comment-511923 for how to fix this. This link tells me how to change some of my share settings but I don't really understand what I'm trying to do as it doesn't move everything off the cache drive so I don't want to proceed removing the cache drive and re-formatting it. Is this a new issue or has it just been introduced with 6.5? The instructions look as if they have been around a while.
March 24, 20188 yr 1 minute ago, PhilDG said: Is this a new issue or has it just been introduced with 6.5? Old issue from 6.3.x Are you on the latest version of FCP?
March 24, 20188 yr 2 minutes ago, PhilDG said: Hi, I've run the upgrade assistant to upgrade from 6.41 to 6.5 and it gives me this message: Issue Found: Cache drive partition doesn't start on sector 64. You will have problems. See here https://lime-technology.com/forums/topic/46802-faq-for-unraid-v6/?tab=comments#comment-511923 for how to fix this. This link tells me how to change some of my share settings but I don't really understand what I'm trying to do as it doesn't move everything off the cache drive so I don't want to proceed removing the cache drive and re-formatting it. Is this a new issue or has it just been introduced with 6.5? The instructions look as if they have been around a while. What the instructions are telling you is how to move the data off the cache by setting the cache preferences on the shares. Then you can reformat the cache disk and move the data back onto the newly formatted cache. Read the instructions carefully.
March 24, 20188 yr 11 minutes ago, PhilDG said: Fix Common Problems? Yes I think so Then do me a favour and post your diagnostics and I'll give you a command to run to see why it's failing here
March 24, 20188 yr Thanks Squid I have just run FCP in extended mode and got some SMB file permission issues but it doesn't report anything about the cache. Here is the diagnostics tower-diagnostics-20180324-2124.zip
March 24, 20188 yr 11 minutes ago, PhilDG said: Thanks Squid I have just run FCP in extended mode and got some SMB file permission issues but it doesn't report anything about the cache. Here is the diagnostics tower-diagnostics-20180324-2124.zip Give me the output of this command from a terminal if you're so inclined: fdisk -l /dev/sdb Edited March 24, 20188 yr by Squid
March 24, 20188 yr 1 minute ago, Squid said: Give me the output of this command from a terminal if you're so inclined: fdisk -l /dev/sdb root@tower:/# fdisk -l /dev/sdb Disk /dev/sdb: 232.9 GiB, 250059350016 bytes, 488397168 sectors Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disklabel type: dos Disk identifier: 0x00000000 Device Boot Start End Sectors Size Id Type /dev/sdb1 63 488397167 488397105 232.9G 83 Linux root@tower:/#
Archived
This topic is now archived and is closed to further replies.