December 28, 20169 yr Hey guys i hope you are well ! I heared of unRAID and watched quite a few videos from linustechtips yt channel and i love it! Before i go into my questions, im also familiar with linux and raid levels, but i got a few question about unRAIDs implementation of performance. Afaik i should use 2 SSDs for raid1 cache disks and a mechanical raid 5 array existing from 5 harddisks and in the midnight everything from cache is moved to mechanical storage. Now to my Questions: So basically im running Windows10 with GPU bypass over Intel VT-D/X but the read speed performance of my Windows10-VM is not from the SSD raid1 or the cache from the SSDs, but from the 4 times more readspeed of 5 mechanial in raid5 right? The SSDs are just a means to an end, which is to aquire very fast write speed over SMB/NAS connections or am i wrong? What happens if i write massive data inside my Windows10-VM, does the SSD -cache work there to? Again im familiar with linux, raid levels, virtualization and other products like proxmox, so please talk geeky if it explains it better. Im very interested in unRAID, because of all the features! Sorry for that much questions in my first post, but im really confused here and i see no clear answer, i hope you can help me, because im planning to update my gaming rig to NAS+gaming, but i need to know the facts first Greetings, jammsen
December 28, 20169 yr There is no Raid-Level in unRAID - except for the Cache-Pool (Raid 1). So what is the Question?
December 28, 20169 yr Community Expert unRAID is not raid. The key thing to understand about the unRAID portion of this OS is there is NO striping. Each disk in the parity array is an independent filesystem. Reading takes place at the speed of an individual disk, and writing is slower since parity must also be calculated and written. This has the advantage of allowing mixed disk sizes in the array, and also limits your losses to only the affected drives in the event you have more failures than parity. It also allows drives that are not being accessed to be spun down. Cache is not part of the parity array. A single cache disk is written and read at the speed of the single drive. Cache pools provide additional capabilities. The default for the cache pool is raid1, which only provides redundancy. Other raid configurations are possible if you have enough drives in the cache pool, including raid10. User shares can span drives. Cache can be used to cache writes to any user share, thus writing at the speed of cache instead of at the slower speed of the parity array. Cached user share writes are moved to the parity array at a scheduled time. Cache can also be used for dockers and VMs, thus avoiding the write speed penalty of the parity array. And dockers and VMs can also access the larger protected storage of the parity array. I think that should pretty much cover the storage aspects of the system. Hope this helps and don't hesitate to ask if anything is unclear or you have any other questions.
December 30, 20169 yr Author Hey guys, thanks a lot for your informations. Is there like a beginners tutorial for new guys on how raidlevels like 5 with parity disks and unRAIDs version is implemented? Do i loose data if i get an hardware error from an array of 5 disks? How do i regenerate lost data in the array? Is it possbile to use a hardware or software raid in front of unRAID (if needed)?
December 30, 20169 yr Community Expert Hey guys, thanks a lot for your informations. Is there like a beginners tutorial for new guys on how raidlevels like 5 with parity disks and unRAIDs version is implemented? Do i loose data if i get an hardware error from an array of 5 disks? How do i regenerate lost data in the array? Is it possbile to use a hardware or software raid in front of unRAID (if needed)? Wiki - unRAID 6 Manual
December 30, 20169 yr Hey guys, thanks a lot for your informations. Is there like a beginners tutorial for new guys on how raidlevels like 5 with parity disks and unRAIDs version is implemented? Do i loose data if i get an hardware error from an array of 5 disks? How do i regenerate lost data in the array? Is it possbile to use a hardware or software raid in front of unRAID (if needed)? Trurl already gave you the link to the manual, but in particular you should look at: http://lime-technology.com/wiki/index.php/UnRAID_6/Overview#Parity-Protected_Array unRAID is a software RAID implementation. Before you worry about other forms of hardware or software RAID it's best to understand what unRAID has to offer. It can protect you hardware errors, regenerate lost data, and perform all the critical functions of hardware RAID, but with a lot more flexibility.
December 31, 20169 yr Author Trurl already gave you the link to the manual, but in particular you should look at: http://lime-technology.com/wiki/index.php/UnRAID_6/Overview#Parity-Protected_Array unRAID is a software RAID implementation. Before you worry about other forms of hardware or software RAID it's best to understand what unRAID has to offer. It can protect you hardware errors, regenerate lost data, and perform all the critical functions of hardware RAID, but with a lot more flexibility. That sounds pretty dope! Thanks a lot, for all your help guys! I will read up on the informations provided byyou. If i dont understand something, i will get back here, hope thats okay.
January 9, 20179 yr Author So after a weekend full hours and hours and tests over tests, i have to say, i love the idea of unRAID, sadly i think the support for skylake based hardware is still missing and i had to roll back to my baremetall windows 10 machine. I've encountered so many errors and bugs and tried one after another to bugfix them with help of the search function of this forum and google, but i run still in situations where nothing even worked. What i did: [*]Image from Windows 10 SSD with Acronis, in case, you know [*]Adding more drives to the system (Before: 1xSSD Samsung 850 PRO 512G + 1x 2TB WD Green 5400rpm disk for just data - Now: 1x SSD 850 PRO 512G, 1x SSD SandSisk 120 Ultra 2, 2x 1TB WD Claviar Black, 1x 2tb WD Green) [*]Added 2nd GPU to the system, now 1x ZOTAC GTX 770 AMP! + 1x ZOTAC GTX 560 Ti AMP! [*]Function check under Windows 10, everything works fine [*]Made a unRAID USB-Stick and changed first gpu to internal (iGPU) [*]Booted unRAID and watched a lot of linus setup videos for the Mainpage disk setup, it really took a while and i redid that like 3 times, because i was not sure on what and where, example: i added both ssds to cache, which made them a raid 1 with about 315G because of 512G and 120G capacity, but i wanted to run my main win10 machine from the 850 PRO, because of the best performance, so i had to remake the whole thing again and added the 850 PRO to data array and changed shares for domains to disk 3 (850 PRO) only (i permit everything else to write on that too) [*]After all this i really felt the slow down while parity generation, so i did some setup stuff and tried to not do disk intense stuff [*]After parity was done i had basically the setup done, with win10 vms and all, of course i managed to not read hyper-v for nvidia gfx cards, so i let it on, which was wrong, the bios was not SeaBios and i was far from bypassing the usb 3 host controller [*]Frist step was to get the damn win10 vm starting, but i got always an iommu group error, because of Skylakes onbaord soundcard stuff, so i changed that to HDMI Audio, then the VM booted and i didnt installed win10 new, i added the acronis image tool and recovered my external baremetal snapshot, which did run a little [*]After a while the windows installation stopped working entirely, it hong but was not dead, if you know what i mean, i even could reach network, via ping on my router by ip, but dns reslove took like forever on a ping heise.de or so, so that wasnt cool [*]I didnt get that fixed, so i tried to get the soundcard running, but men, that was a nightmare, i added stubs to te wrong startimage, after a while i realized that and i was getting the usb plug and play running, but no where near the soundcard, i read 20 minutes on this forum to get it running, via stubs or blacklisting kernel modules, but it never did run [*]After a while i thought, i could have damaged my already used win10 installation and thought, force stop this for now and make a new win10 installation, after about 8 minutes the new system was done, but the same weird networking and reaction on everything stuff was there too and sound never did work for onboard audio, just usb Logitech 933 headset worked, but thats not an option for me [*]So i stopped that for now and restored my win10 image, for now! Im not ready to quit yet [*]My goal is: the have an NAS underground and 2 Gaming Rigs, the 1st with like a quadcore and the GTX770 on the 850 PRO a, the 2nd with like 3 cores and the gtx 560 ti on the SanDisk Cache Disk for friend visit gaming, with separate mouse and keyboard Hope it was an entertaining read for my first time, i really hope you guys can help me get this stuff running. Here are some Shell-Outputs: i still have lspci: 00:00.0 Host bridge: Intel Corporation Skylake Host Bridge/DRAM Registers (rev 07) 00:01.0 PCI bridge: Intel Corporation Skylake PCIe Controller (x16) (rev 07) 00:01.1 PCI bridge: Intel Corporation Skylake PCIe Controller (x8) (rev 07) 00:02.0 VGA compatible controller: Intel Corporation HD Graphics 530 (rev 06) 00:08.0 System peripheral: Intel Corporation Skylake Gaussian Mixture Model 00:14.0 USB controller: Intel Corporation Sunrise Point-H USB 3.0 xHCI Controller (rev 31) 00:14.2 Signal processing controller: Intel Corporation Sunrise Point-H Thermal subsystem (rev 31) 00:15.0 Signal processing controller: Intel Corporation Sunrise Point-H Serial IO I2C Controller #0 (rev 31) 00:15.1 Signal processing controller: Intel Corporation Sunrise Point-H Serial IO I2C Controller #1 (rev 31) 00:16.0 Communication controller: Intel Corporation Sunrise Point-H CSME HECI #1 (rev 31) 00:17.0 SATA controller: Intel Corporation Sunrise Point-H SATA controller [AHCI mode] (rev 31) 00:1c.0 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #1 (rev f1) 00:1c.2 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #3 (rev f1) 00:1d.0 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #9 (rev f1) 00:1d.2 PCI bridge: Intel Corporation Sunrise Point-H PCI Express Root Port #11 (rev f1) 00:1e.0 Signal processing controller: Intel Corporation Sunrise Point-H Serial IO UART #0 (rev 31) 00:1f.0 ISA bridge: Intel Corporation Sunrise Point-H LPC Controller (rev 31) 00:1f.2 Memory controller: Intel Corporation Sunrise Point-H PMC (rev 31) 00:1f.3 Audio device: Intel Corporation Sunrise Point-H HD Audio (rev 31) 00:1f.4 SMBus: Intel Corporation Sunrise Point-H SMBus (rev 31) 00:1f.6 Ethernet controller: Intel Corporation Ethernet Connection (2) I219-V (rev 31) 01:00.0 VGA compatible controller: NVIDIA Corporation GK104 [GeForce GTX 770] (rev a1) 01:00.1 Audio device: NVIDIA Corporation GK104 HDMI Audio Controller (rev a1) 02:00.0 VGA compatible controller: NVIDIA Corporation GF114 [GeForce GTX 560 Ti] (rev a1) 02:00.1 Audio device: NVIDIA Corporation GF114 HDMI Audio Controller (rev a1) 04:00.0 USB controller: ASMedia Technology Inc. ASM1142 USB 3.1 Host Controller 06:00.0 PCI bridge: ASMedia Technology Inc. ASM1083/1085 PCIe to PCI Bridge (rev 03) lspci -n: 00:00.0 0600: 8086:191f (rev 07) 00:01.0 0604: 8086:1901 (rev 07) 00:01.1 0604: 8086:1905 (rev 07) 00:02.0 0300: 8086:1912 (rev 06) 00:08.0 0880: 8086:1911 00:14.0 0c03: 8086:a12f (rev 31) 00:14.2 1180: 8086:a131 (rev 31) 00:15.0 1180: 8086:a160 (rev 31) 00:15.1 1180: 8086:a161 (rev 31) 00:16.0 0780: 8086:a13a (rev 31) 00:17.0 0106: 8086:a102 (rev 31) 00:1c.0 0604: 8086:a110 (rev f1) 00:1c.2 0604: 8086:a112 (rev f1) 00:1d.0 0604: 8086:a118 (rev f1) 00:1d.2 0604: 8086:a11a (rev f1) 00:1e.0 1180: 8086:a127 (rev 31) 00:1f.0 0601: 8086:a145 (rev 31) 00:1f.2 0580: 8086:a121 (rev 31) 00:1f.3 0403: 8086:a170 (rev 31) 00:1f.4 0c05: 8086:a123 (rev 31) 00:1f.6 0200: 8086:15b8 (rev 31) 01:00.0 0300: 10de:1184 (rev a1) 01:00.1 0403: 10de:0e0a (rev a1) 02:00.0 0300: 10de:1200 (rev a1) 02:00.1 0403: 10de:0e0c (rev a1) 04:00.0 0c03: 1b21:1242 06:00.0 0604: 1b21:1080 (rev 03) Audio Devices: root@Tower:~# lspci | grep Audio 00:1f.3 Audio device: Intel Corporation Sunrise Point-H HD Audio (rev 31) 01:00.1 Audio device: NVIDIA Corporation GK104 HDMI Audio Controller (rev a1) 02:00.1 Audio device: NVIDIA Corporation GF114 HDMI Audio Controller (rev a1) Ignore Stuff for Sysconfig: pci-stub.ids=8086:a12f modprobe.blacklist=i2c_i801,i2c_smbus Ifconfig: root@Tower:~# ifconfig bond0: flags=5443<UP,BROADCAST,RUNNING,PROMISC,MASTER,MULTICAST> mtu 1500 ether d8:cb:8a:c3:28:f5 txqueuelen 1000 (Ethernet) RX packets 89075 bytes 37072435 (35.3 MiB) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 113989 bytes 77112658 (73.5 MiB) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 br0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500 inet 192.168.178.33 netmask 255.255.255.0 broadcast 192.168.178.255 ether d8:cb:8a:c3:28:f5 txqueuelen 1000 (Ethernet) RX packets 883674 bytes 88527113 (84.4 MiB) RX errors 0 dropped 1260 overruns 0 frame 0 TX packets 693585 bytes 30422040045 (28.3 GiB) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 docker0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500 inet 172.17.0.1 netmask 255.255.0.0 broadcast 0.0.0.0 ether 02:42:e4:0d:18:2b txqueuelen 0 (Ethernet) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 0 bytes 0 (0.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 eth0: flags=6211<UP,BROADCAST,RUNNING,SLAVE,MULTICAST> mtu 1500 ether d8:cb:8a:c3:28:f5 txqueuelen 1000 (Ethernet) RX packets 89075 bytes 37072435 (35.3 MiB) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 113989 bytes 77112658 (73.5 MiB) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 device interrupt 16 memory 0xdf200000-df220000 lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536 inet 127.0.0.1 netmask 255.255.255.255 loop txqueuelen 1 (Local Loopback) RX packets 2 bytes 140 (140.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 2 bytes 140 (140.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 virbr0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500 inet 192.168.122.1 netmask 255.255.255.0 broadcast 192.168.122.255 ether 52:54:00:0a:e6:60 txqueuelen 1000 (Ethernet) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 0 bytes 0 (0.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
January 9, 20179 yr Community Expert I will let others address any VM questions you may have, but I thought I would comment on this Now: 1x SSD 850 PRO 512G, 1x SSD SandSisk 120 Ultra 2 You don't say exactly how you are using these SSDs. If you have both of them in a btrfs cache pool with the default settings of raid1 then you are only getting 120GB of storage. btrfs disk usage calculator
January 9, 20179 yr Author You don't say exactly how you are using these SSDs. If you have both of them in a btrfs cache pool with the default settings of raid1 then you are only getting 120GB of storage. I wrote: ... which made them a raid 1 with about 315G because of 512G and 120G capacity, but i wanted to run my main win10 machine from the 850 PRO, because of the best performance, so i had to remake the whole thing again and added the 850 PRO to data array and changed shares for domains to disk 3 (850 PRO) only (i permit everything else to write on that too) ... Which means, in the end i used the 512G 850 Pro as a array member, only for domain purposes and the other 120G as Cache, which in my case is fairly enough
January 9, 20179 yr Author Changed topic name / title from: "Starter questions about raidlevel implementation and performance" to: "Changed title: Need help with unRAID i7 Skylake Arch" That fits now more the context.
January 9, 20179 yr Community Expert You don't say exactly how you are using these SSDs. If you have both of them in a btrfs cache pool with the default settings of raid1 then you are only getting 120GB of storage. I wrote: ... which made them a raid 1 with about 315G because of 512G and 120G capacity, but i wanted to run my main win10 machine from the 850 PRO, because of the best performance, so i had to remake the whole thing again and added the 850 PRO to data array and changed shares for domains to disk 3 (850 PRO) only (i permit everything else to write on that too) ... Which means, in the end i used the 512G 850 Pro as a array member, only for domain purposes and the other 120G as Cache, which in my case is fairly enough 315G is incorrect. Maybe that is what the webUI reported, but still wrong. Since you aren't doing this now I guess it doesn't matter but for future reference you might check the link I gave. There is still some controversy regarding SSDs in the array. There has been some concern that SSDs built-in storage management (TRIM) could lead to invalid parity. You might do a search of the forum about this. See How to Search in my sig.
January 10, 20179 yr Author Thanks Trurl, i already did, the calculator is good, while im not sure how that would really work i could try that yes. On the part of ssd in array i will research that, thanks for the tip. Do you got anything of help, on the part of having really slow dns/internet inside the VMs and barely reaction, while in my opinion the systems had no load what so ever?! On-Board sound in my main VM would be nice too
Archived
This topic is now archived and is closed to further replies.