trurl Posted May 16, 2019 Share Posted May 16, 2019 10 minutes ago, nlucier said: running out of memory (currently have 2G but working to rectify) ... When I try to update the OS to 6.7.0, this is the result in the script window: You won't be able to update in the webUI with so little RAM. You will have to do a manual update, or just wait for your memory upgrade. Quote Link to comment
JonathanM Posted May 16, 2019 Share Posted May 16, 2019 One of my test servers has 2GB of RAM, and I was able to successfully update from the webUI, but not if the array was already started. I suspect it's right on the edge, so doing the update after a fresh reboot in safe mode with the array NOT autostarted may possibly succeed. YMMV. It may also help to disable docker and VM services if they are not already disabled on a machine with that little power. Quote Link to comment
warwon Posted May 17, 2019 Share Posted May 17, 2019 From my viewpoint no reason not to roll back as 6.7 is broken. There are a ton of people reporting videos freezing using dockers running Plex and direct network play videos. As of now it unstable, it went from being useable to not. Simple, roll back Quote Link to comment
kizer Posted May 17, 2019 Share Posted May 17, 2019 2 hours ago, warwon said: From my viewpoint no reason not to roll back as 6.7 is broken. There are a ton of people reporting videos freezing using dockers running Plex and direct network play videos. As of now it unstable, it went from being useable to not. Simple, roll back Zero issues here. Quote Link to comment
itimpi Posted May 17, 2019 Share Posted May 17, 2019 2 hours ago, warwon said: From my viewpoint no reason not to roll back as 6.7 is broken. There are a ton of people reporting videos freezing using dockers running Plex and direct network play videos. As of now it unstable, it went from being useable to not. Simple, roll back There are also a very large number of people reporting no problems at all with 6.7 so rushing to roll back can be premature. The problem with rolling back is that you may end up stuck on an old release unnecessarily. 1 Quote Link to comment
JorgeB Posted May 17, 2019 Share Posted May 17, 2019 9 hours ago, [email protected] said: Re-upgraded to 6.7.0 and experienced the same issues as before. None of my SATA drives connected to the MoBo were visible in Unraid after the update. I've included the diagnostics zip from after the upgrade. Hopefully this can be of some help. I will be staying on 6.6.7 for now. There is a known issue with the 9230 Marvell controller and the newer kernels, you can usually get around it with this, but better yet, use the Intel onboard ports instead, since they are all unused, the Marvell controller (first 4 white SATA ports) on those boards is known to drop disks. May 16 17:14:57 Basestar kernel: ata3: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata2: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata5: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata4: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata1: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata6: SATA link down (SStatus 0 SControl 300) Quote Link to comment
karldonteljames Posted May 17, 2019 Share Posted May 17, 2019 Morning All. I've just upgraded to 6.7.0 and i can see that some people are having problems with missing drives (this is also the issue i'm having.) I'm not sure if this is down to a marvel controller or not. In the linked thread above it says about modifying a sylinux file, can I do this from the webui terminal, if so where are the files located that I need to modify. I've attached a copy of my diagnostic report. Hopefully someone will be able to help. maximus-diagnostics-20190517-0826.zip Quote Link to comment
Zonediver Posted May 17, 2019 Share Posted May 17, 2019 (edited) Upgraded from 6.6.7 and no problems so far - thanks for the good work 😉 Edited May 17, 2019 by Zonediver Quote Link to comment
bonienl Posted May 17, 2019 Share Posted May 17, 2019 6 minutes ago, karldonteljames said: Morning All. I've just upgraded to 6.7.0 and i can see that some people are having problems with missing drives (this is also the issue i'm having.) I'm not sure if this is down to a marvel controller or not. In the linked thread above it says about modifying a sylinux file, can I do this from the webui terminal, if so where are the files located that I need to modify. I've attached a copy of my diagnostic report. Hopefully someone will be able to help. maximus-diagnostics-20190517-0826.zip 97.42 kB · 0 downloads Yes, you have a Marvell controller type (9230) which is affected 0a:00.0 SATA controller [0106]: Marvell Technology Group Ltd. 88SE9230 PCIe SATA 6Gb/s Controller [1b4b:9230] (rev 11) Subsystem: Marvell Technology Group Ltd. 88SE9230 PCIe SATA 6Gb/s Controller [1b4b:9230] Go to Main -> Flash -> Syslinux Configuration Change the "append" line of the default (green) section to for Intel processors append iommu=pt initrd=/bzroot for AMD processors append amd_iommu=pt initrd=/bzroot Apply and reboot your system 1 1 Quote Link to comment
karldonteljames Posted May 17, 2019 Share Posted May 17, 2019 1 minute ago, bonienl said: Yes, you have a Marvell controller type (9230) which is affected 0a:00.0 SATA controller [0106]: Marvell Technology Group Ltd. 88SE9230 PCIe SATA 6Gb/s Controller [1b4b:9230] (rev 11) Subsystem: Marvell Technology Group Ltd. 88SE9230 PCIe SATA 6Gb/s Controller [1b4b:9230] Go to Main -> Flash -> Syslinux Configuration Change the "append" line of the default (green) section to for Intel processors append iommu=pt initrd=/bzroot for AMD processors append amd_iommu=pt initrd=/bzroot Apply and reboot your system Thanks for the help. I rolled back and rebooted, but the server is yet to come back online. As soon as it does, I'll do that. Will I need to do that after each unraid update? 1 Quote Link to comment
itimpi Posted May 17, 2019 Share Posted May 17, 2019 1 minute ago, karldonteljames said: Thanks for the help. I rolled back and rebooted, but the server is yet to come back online. As soon as it does, I'll do that. Will I need to do that after each unraid update? No. You would only need to redo this if you reformatted the USB stick for some reason to put a ‘virgin’ Unraid install on it. The syslinux settings do not get changed by simply applying an update to Unraid. Quote Link to comment
cybrnook Posted May 17, 2019 Share Posted May 17, 2019 No problems on 6.7.0, great work everyone! Quote Link to comment
darianf Posted May 17, 2019 Share Posted May 17, 2019 I updated 2 towers from 6.6.7 to 6.7.0 without any issue. 1 optiplex 9010 and 1 asus x99-m ws (passthrough 1 ssd, 2 nvidia 10xx cards and 1 usb controller) no marvel controllers thou... Thank you for another great update! Quote Link to comment
Rajenator Posted May 17, 2019 Share Posted May 17, 2019 (edited) Running UNRAID 6.6.7 on a I7-4930K/EVGA X79 Dark setup...mobo has both chipset and MARVELL SATA controllers. Same issue with the MARVELL controller for me when attempted an upgrade from 6.6.7 to 6.7.0 versions (stable)....array stopped, missing disks, start button greyed-out. Once rolled back from 6.7.0 to 6.6.7 (and rebooted), the array started up, everything was fine, and all plugins and drives were found and were working. Any ideas for if/when this would be addressed in a future release would be greatly appreciated. While this is only my test/Lab server (I am learning UNRAID, VMs, Docker containers, etc.), I am sure that others running systems with MARVELL SATA controllers are managing data that potentially would be a bit more important (i.e. production) than mine. Please and Thank You, Raj Edited May 17, 2019 by Rajenator Quote Link to comment
itimpi Posted May 17, 2019 Share Posted May 17, 2019 (edited) 32 minutes ago, Rajenator said: Running UNRAID 6.6.7 on a I7-4930K/EVGA X79 Dark setup...mobo has both chipset and MARVELL SATA controllers. Same issue with the MARVELL controller for me when attempted an upgrade from 6.6.7 to 6.7.0 versions (stable)....array stopped, missing disks, start button greyed-out. Once rolled back from 6.7.0 to 6.6.7 (and rebooted), the array started up, everything was fine, and all plugins and drives were found and were working. Any ideas for if/when this would be addressed in a future release would be greatly appreciated. While this is only my test/Lab server (I am learning UNRAID, VMs, Docker containers, etc.), I am sure that others running systems with MARVELL SATA controllers are managing data that potentially would be a bit more important (i.e. production) than mine. Please and Thank You, Raj Have you tried the change to the syslinux configuration that seems to have fixed this for many users? Since nobody knows exactly why this is happening in the first place I would not think an ETA for a fix (if one can be found) is going to be given. Edited May 17, 2019 by itimpi 1 Quote Link to comment
JorgeB Posted May 17, 2019 Share Posted May 17, 2019 55 minutes ago, Rajenator said: Running UNRAID 6.6.7 on a I7-4930K/EVGA X79 Dark setup...mobo has both chipset and MARVELL SATA controllers. Same issue with the MARVELL controller for me when attempted an upgrade from 6.6.7 to 6.7.0 versions (stable)....array stopped, missing disks, start button greyed-out. See just 4 or 5 post above for a possible solution. 1 Quote Link to comment
Kujo Posted May 17, 2019 Share Posted May 17, 2019 (edited) SMB broken with OpenELEC 9.02 after upgrading to 6.7. No issues connecting to the Unraid SMB shares on other devices. (i.e Windows, Android). Downgrading back to 6.6.7 workaround for now. Edited May 17, 2019 by Kujo Quote Link to comment
Rajenator Posted May 17, 2019 Share Posted May 17, 2019 3 hours ago, johnnie.black said: See just 4 or 5 post above for a possible solution. Thanks very much to ALL who responded to my first ever post here. The fix worked fine post-update to 6.7.0, and the array shows as complete. Best Regards, Raj Quote Link to comment
[email protected] Posted May 17, 2019 Share Posted May 17, 2019 11 hours ago, johnnie.black said: There is a known issue with the 9230 Marvell controller and the newer kernels, you can usually get around it with this, but better yet, use the Intel onboard ports instead, since they are all unused, the Marvell controller (first 4 white SATA ports) on those boards is known to drop disks. May 16 17:14:57 Basestar kernel: ata3: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata2: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata5: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata4: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata1: SATA link down (SStatus 0 SControl 300) May 16 17:14:57 Basestar kernel: ata6: SATA link down (SStatus 0 SControl 300) Thank you for the quick reply. I’m not surprised that there is an issue with the Marvell Controller and the Kernel. When I have some downtime to take the server offline I will move those 4 drives to the Intel SATA ports and try again. Quote Link to comment
landS Posted May 17, 2019 Share Posted May 17, 2019 (edited) good day @limetech Given that under my Flash/Syslinux config I have: append iommu=pt vfio-pci.ids=1b21:0612,1b21:1242 isolcpus=2-4,8-10 initrd=/bzroot *Question 1*Do I need to remove the vfio-pic from the Flash/Syslinxu and instead use the new flashconfig/vfio-pci.cfg BIND=0c:03.0 0d:00.0 03:00.0 Where IOMMU Group 39 [1b21:1182] 0c:03.0 PCI bridge: ASMedia Technology Inc. Device 1182 Group 39 [1b21:0612] 0d:00.0 SATA controller: ASMedia Technology Inc. ASM1062 Serial ATA Controller group 31 [1b21:1242] 03:00.0 USB controller: ASMedia Technology Inc. ASM1142 USB 3.1 Host Controller *Question 2*Should my 2 GPUs need to be added to this as well? group 32 [10de:1c02] 04:00.0 VGA compatible controller: NVIDIA Corporation GP106 [GeForce GTX 1060 3GB] group 32 [10de:10f1] 04:00.1 Audio device: NVIDIA Corporation GP106 High Definition Audio Controller group 30 [10de:1b80] 02:00.0 VGA compatible controller: NVIDIA Corporation GP104 [GeForce GTX 1080] group 30 [10de:10f0] 02:00.1 Audio device: NVIDIA Corporation GP104 High Definition Audio Controller BIND=0c:03.0 0d:00.0 03:00.0 04:00.0 04:00.1 02:00.0 02:00.1 *Question 3* Should I remove isolcpus=2-4,8-10 given that I have pinning selected via the GUI Thanks! BTW, initial update looks VERY good Supermicro X10SRA-F Windows 10 VM with 1080 gpu and BT adapater passed through for SteamLink Streaming Windows 10 VM with 1060 gpu, usb adapter, ODD sata controller passed through for production Crashplan docker Edited May 17, 2019 by landS Quote Link to comment
limetech Posted May 17, 2019 Author Share Posted May 17, 2019 18 minutes ago, landS said: good day @limetech Given that under my Flash/Syslinux config I have: append iommu=pt vfio-pci.ids=1b21:0612,1b21:1242 isolcpus=2-4,8-10 initrd=/bzroot *Question 1*Do I need to remove the vfio-pic from the Flash/Syslinxu and instead use the new flashconfig/vfio-pci.cfg BIND=0c:03.0 0d:00.0 03:00.0 Where IOMMU Group 39 [1b21:1182] 0c:03.0 PCI bridge: ASMedia Technology Inc. Device 1182 Group 39 [1b21:0612] 0d:00.0 SATA controller: ASMedia Technology Inc. ASM1062 Serial ATA Controller group 31 [1b21:1242] 03:00.0 USB controller: ASMedia Technology Inc. ASM1142 USB 3.1 Host Controller *Question 2*Should my 2 GPUs need to be added to this as well? group 32 [10de:1c02] 04:00.0 VGA compatible controller: NVIDIA Corporation GP106 [GeForce GTX 1060 3GB] group 32 [10de:10f1] 04:00.1 Audio device: NVIDIA Corporation GP106 High Definition Audio Controller group 30 [10de:1b80] 02:00.0 VGA compatible controller: NVIDIA Corporation GP104 [GeForce GTX 1080] group 30 [10de:10f0] 02:00.1 Audio device: NVIDIA Corporation GP104 High Definition Audio Controller BIND=0c:03.0 0d:00.0 03:00.0 04:00.0 04:00.1 02:00.0 02:00.1 *Question 3* Should I remove isolcpus=2-4,8-10 given that I have pinning selected via the GUI Thanks! BTW, initial update looks VERY good Supermicro X10SRA-F Windows 10 VM with 1080 gpu and BT adapater passed through for SteamLink Streaming Windows 10 VM with 1060 gpu, usb adapter, ODD sata controller passed through for production Crashplan docker No, don't need to make any changes if all is working as it should. 1 Quote Link to comment
landS Posted May 18, 2019 Share Posted May 18, 2019 22 minutes ago, limetech said: No, don't need to make any changes if all is working as it should. Awesome, thanks again. The Gui polish is nice BTW Quote Link to comment
Pro-289 Posted May 18, 2019 Share Posted May 18, 2019 guinea pig 22 hours ago, itimpi said: There are also a very large number of people reporting no problems at all with 6.7 so rushing to roll back can be premature. The problem with rolling back is that you may end up stuck on an old release unnecessarily. Using an old release may not be all that bad. An old release is just a previous new release. Sometimes updating to the latest and greatest causes the guinea pig effect. [cough]Windows 10[cough] Quote Link to comment
John_M Posted May 18, 2019 Share Posted May 18, 2019 3 hours ago, Pro-289 said: Using an old release may not be all that bad. An old release is just a previous new release. Sometimes updating to the latest and greatest causes the guinea pig effect. [cough]Windows 10[cough] That's true and if you're living in isolation it does't really matter if you don't upgrade. But the reality is that it's a big bad world out there that has become bigger and badder since that previous release was new. But you're not living in isolation because you posted here, which means you're connected, which means you're vulnerable. You may not need the new features but you certainly need the security updates. Quote Link to comment
Keyko Posted May 18, 2019 Share Posted May 18, 2019 (edited) What about "all ssd array"? The last release had a problem and with parity check CRC Error appeared in SMART. (CRC Error grew on parity ssd ) Edited May 18, 2019 by Keyko Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.