luca2
-
Posts
659 -
Joined
-
Last visited
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Posts posted by luca2
-
-
My 2 cents: as stated before if you need fps you need BM. It depends what games you play. And if you need high fps or not.
I do have tested BM VS vm on Un raid in Sim racing games and no VR (4k TV). It depends on the game but some do suffer and not provide smooth gameplay. You notice it more if you are playing online.
Once I switched back to BM everything is fine again. Of course if I would run a 2080ti instead of a vega64 I would get better fps.
Also WC is a must. I even recommend you place you rads out of the case!
Ii naked a huge difference. I placed them outside the window!
Enviado desde mi HMA-L29 mediante Tapatalk-
1
-
-
ok, thx for your help. I will re-run a preclear cycle to check again for those 2 parameters.
-
Hi!,
I need to upgrade my actual parity drive (3TB) to this new ST8000DM004 8TB drive.
I would appreciate if anyone can check this smart report and confirm to me the disk is ok.
It seems obvious since the report I think does not show any error, but the new disk did not finish 1 full preclear cycle. I am still waiting to get some feedback on the preclear plugin support forum, but if the disk is ok then I would proceed to upgrade my parity now.
Rgds
-
Hi! Sorry for posting in the general support section.
Just got an error while preclearing a new 8TB drive. Can someone suggest me what to do?
I dont see any explicit error. It just could not finish reading it and exited.
EDIT: I ran a smart test and it shows the disk is ok. Since I want to replace the parity disk with this new disk, can I do that or should I run any further test to check the health of the disk?
Rgds
-
Thx for support!
Had to do some manual editing though. The gpu audio was not erased by the gui.
Rgds
-
Hi!
Just got an error while preclearing a new 8TB drive. Can someone suggest me what to do? Maybe try another time again.
I dont see any explicit error. It just could not finish reading it and exited.
Rgds
-
Hi,
I do have a vm with my gpu passthroughed succesfully. But I need to temporaly put the gpu out. How should I proceed (change) in my xml to revert the gpu passthrough to a normal vm with just the kvm graphics?
I took the gpu off the MB, and now the xml GUI does not show it (obviously). I guess I need to make a minor change in my xml to back my vm back to life without the gpu. Just dont want to makle a bad move. Any help is appreciated.
I attach my xml with passthrough.
-
Hi,
I installed my dad (user with no expertise in IT) unraid. He has been transferring all his files (pics, docs, etc) to several shares in unraid. Recently he called me and I had a look and found one of his disks was showing errors.
At this moment I upgraded his version from 6.5.X to the latest one. No errors showed up.
So then I replaced disk2 with another one and let it rebuild.
After that his array shows no info in any share and also some weird info on the general tab of the GUI. In the pic "array" you can see disk 1 and disk2(both are 1TB drives) are half full. But if you look at the pic "shares" all of them show 2.02TB free space.
Fix common problems (als I attach a pic) shows read errors on all disks! Myself do have unraid running for years and never saw this.
Also, I cannot access or peform a smart anaylisis of any disk (I attach a report).
There must happened something but I can´t tell since I am not next to dad´s server.
Maybe someone can help me out to try to recover the data in the shares.
Thx in advance.
Rgds.
towerpapa-diagnostics-20200114-1957.zip towerpapa-smart-20200114-2002.zip
-
On 12/17/2019 at 11:02 PM, Skitals said:
No experience with aida64, but here are my results (aida64 v6.20.5300)
Memory Read: 53295 MB/s
Memory Write: 44206 MB/s
Memory Copy: 51345 MB/s
Memory Latency: 87.4 ns
Edit: I had noticed high cpu usage while the guest report 1%, changing this from no to yes helped a bit and brough latency "down" to 83.3 ns. Still seems a bit high, yes?
Edit 2: Okay, I got memory latency "down" to 79.0ns AND reduced my idle cpu usage to practically 0% by making these two changes:
Switched the usb controller to 3.0 nec.
Turned hpet back OFF and added these hyperv flags:
Here are the new results:
Memory Read: 56165 MB/s
Memory Write: 44619 MB/s
Memory Copy: 52043 MB/s
Memory Latency: 79.0 ns
Thx a lot! Sorry for late replay but you solved one of my doubts about vm vs baremetal on ryzen. As you can see in my pic (it is my baremetal) the latency I achieve is 66.2ns, so not that far from yours. Memory reads is similar, you have way better WRITES, and similar copy results.
Great info thx!
-
22 hours ago, Skitals said:
Just ran 3dmark Time Spy, here are my refults: https://www.3dmark.com/3dm/41799540
Gigabyte 5700 XT Gaming OC, Ryzen 3900X (passing all but first 2 cores), 16 out of 32GB ram, Windows installed on a 1TB Gen4 m.2 NVME (passing through nvme controller, can also boot Win10 bare metal). Machine type set to Q35-4.0.1. I think my results are pretty good! Getting better numbers in a VM than reviews for this specific vendor card.
Edit: Adrenalin 2020 drivers came out, seeing a nice bump in performance: http://www.3dmark.com/spy/9630130
3dmark score 9252 -> 9761
Graphics score 8986 -> 9533
Hi, wondering if you compared it to your actual barebone setup? I tried it myself with a 3600 and baremetal is more powerful...I guess it depends on the cpu raw power.
Can you run an aida64 ram bench (win 10 VM)?
Rgds
-
On 4/24/2019 at 5:28 AM, rollieindc said:
I guess I need to add myself to the "No Love for the GT610" community.
For me - working with unRAID 6.6.7 on a DELL T310 (with Intel VT-d confirmed) with either SeaBOIS or OVMF, with CPU pass-through or emulated, and building a Windows 10/64bit image VM with the nVidia GT610 in it's own IOMMU group (13) is giving me the same headache in video passthrough as well. Note - I am using TeamViewer (v.14). And I followed the GPU ROM BIOS edits that SpaceInvaderOne had suggested, made sure that there was no added header - and I still can't get the nVidia Card to work reliably, after multiple attempts. I have other VMs that are working fine with TeamViewer, and with VNC/QXL, but nothing seems to be working reliably with the nVidia GT610. I might try one last shot at it with Windows 7/64, but not holding my breath for that one either.
After three nights, I think I am just going to find another video card to try.
did you try this?: see https://wiki.archlinux.org/index.php/PCI_passthrough_via_OVMF
"Error 43: Driver failed to load" on Nvidia GPUs passed to Windows VMs
Note:
This may also fix SYSTEM_THREAD_EXCEPTION_NOT_HANDLED boot crashes related to Nvidia drivers.
This may also fix problems under linux guests.
Since version 337.88, Nvidia drivers on Windows check if an hypervisor is running and fail if it detects one, which results in an Error 43 in the Windows device manager. Starting with QEMU 2.5.0 and libvirt 1.3.3, the vendor_id for the hypervisor can be spoofed, which is enough to fool the Nvidia drivers into loading anyway. All one must do is add hv_vendor_id=1234567890ab to the cpu parameters in their QEMU command line, or by adding the following line to their libvirt domain configuration. The vendor_id can be whatever you prefer, however it is important to note that the vendor_id must be exactly 12 characters long, per the libvirt documentation [2]:
-
1
-
-
CPU pinning with isolcpus ( https://wiki.archlinux.org/index.php/PCI_passthrough_via_OVMF)
Alternatively, make sure that you have isolated CPUs properly. In this example, let us assume you are using CPUs 4-7. Use the kernel parameters isolcpus nohz_full rcu_nocbs to completely isolate the CPUs from the kernel. For example:
isolcpus=4-7 nohz_full=4-7 rcu_nocbs=4-7Then, run qemu-system-x86_64 with taskset and chrt:
# chrt -r 1 taskset -c 4-7 qemu-system-x86_64 ...The chrt command will ensure that the task scheduler will round-robin distribute work (otherwise it will all stay on the first cpu).
Is it necessary to do this?
-
mode='host-passthrough' check='none'> sockets='1' cores='4' threads='2'/> policy='require' name='topoext'/>
Topoext is automatically enabled in guests if the host supports it. Adding it to an AMD host that doesn't support it doesn't force it on. The CoreInfo utility still showed no hyper-threading.
Hi, anyone on ryzen 3000 tried to activate hyper-threading (as jbarlett mentions) and checked for impact on performance? Supposedly it is deactivated since qemu 3. 1. -
Hi, I just bought a ryzen 3600. I am running a test server (new build with latest rc 4.8.6). According to amd this chip has only 1 Core chiplet with a 4 core+4core CCX design (some cores deactivated).
I checked yesterday at how unraid shows them (attached pic). There are 6 pairs.
I read about improving performance of Threadrippers with more cores, but want to ask how would be the optimal pinning for this cpu since I am not sure how to interpret how are the cores shown in Unraid. Maybe some can point me out in the right direction.
Rgds
-
Hi,
Is there an equivalent command in linux to this: ipconfig /flushdns.
Also, where does unraid store the dns?
Rgds.
-
Excellent!. Thx for the info. This is what I was looking for.
Rgds
-
Hi, sorry for late replay, but I wanted to confirm Unraid works as it should.
I recreated the scenario and could not log in. I guess I have added the share while installing the vm, and at this moment I must have used my admin password. Accidentally I let windows save it. Although I desconnected the share later, when the other user tried to connect the share later I guess my password was still there.
I forgot to ask if there is a way to hide the complete published shares that are visible when you try to add a network share? Although you cannot access them you can see them.
Rgds
-
Thx for your help.
I tried last one and it keeps it simple to do a quick scan.
I investigated further and found Nmap. It has a steeper learning curve. But definitely worth a try.
Rgds -
Any test I can perform myself. For example running a test battery from a friend's house or something similar?
-
Hi,
Can anyone let me know how can I check for open ports of my dockers/vms from outside my lan?
Rgds
-
Has anyone tried this new option: nextDNS?
Here is the Reddit link I wrote about:
The user Brodie7838 mentions some problems of his pihole sometimes.
Also, how can I check if my DNS ports are closed effectively by the router firewall?
-
Hi,
in a VM (windows10) that I only granted read/write privileges to access only one share I accidentally find out this.
The share created acts as a data hdd for this VM only.
It works as it should, but if for example I do a screenshot and then try to save it, at the beginning the share is not shown, but I can do a "create an access to a network share". At this moment I can access any share!!!!
What should I check to avoid this?
Rgds
-
Hi, I only opened tcp 445 and it works. Do I need to open also the other ports. I want to avoid as many open ports as possible,
Rgds
-
Hi, thx for all the input. Really interesting stuff.
The more I look into it the more info comes out.
Will definitely have a look at the test. Didn't know about it.
Yesterday I installed opendns free (no registration) and setup my router to do DNS caching and avoid any clients to change their DNS.
I did ping opendns and compare the results quickly vs pinging Google DNS. Google ones are faster but it's ok for me. But will run the test you provided.
Regarding pihole this morning read an article posted on Reddit that mentioned pihole did not have a DNS list updated as it had before. Tbh I have no idea if it is like this or not.
I started reading about DNS and since everything is new I can't really judge anything. I can't tell if opennic is trustfully, etc.....
Only thing I tested and I can speak about is blocking porn with opendns. It works! It was one of my goals. Will try further the registration option opendns offers since you can add more filtering. But will also look at pihole.
Thx for your opinions!
Move data from 3x2Tb drives to 1x8Tb
in General Support
Posted
Hi, running latest version.
Just upgraded my parity to 8Tb.
How should I move the data from 3 drives to 1 bigger size disk?
Enviado desde mi HMA-L29 mediante Tapatalk