Jump to content

ghost82

Members
 View Profile  See their activity

  • Posts

    2,726

  • Joined

  • Last visited

  • Days Won

    19

 Content Type 

Everything posted by ghost82

  1. ghost82
    Your problem is related to dns you are using: your actual dns isn't able to resolve that hostname, change dns.
  2. ghost82
    Why don't you dump your own vbios? SpaceInvaderOne explain how to do it, it's very simple and you can do it from unraid itself.
  3. ghost82
    Sure, but this doesn't mean it's an unraid bug. New code can reveal bugs in third party extensions, and this is what seems to happen, just my 2 cents.
  4. ghost82
    Hi all, this happens on both 6.8.0 and 6.8.1 (can't test previous versions). Bug replication: 1- Have a virtual machine (mine is osx vm with the attached customized xml template - oskey is removed accordingly to forum rules); I'm specifying topology as a qemu:arg value with argument smp, so there's no <topology> line in the xml. All cores are pinned to this vm (16 cores); there's no cpu isolation. 2- Unraid webUI: Settings-->CPU pinning--> CPU pinning VM: all cores are selected: 3- Deselect some cores (in my example cores 0-7 are deselected): 4- Click apply: error "missing sockets attribute in cpu topology" pop up: 5- Click OK 6- in Unraid WebUI go to VM: you can see that the VM template has been deleted (luckily disks are in their places) Issue description: Unraid should popup an error window or better be aware that the topology can be specified by qemu:arg (or not specified at all), but absolutely not deleting the vm template. Take also into account that the <topology> line or the smp+topology are not a must for an osx vm, as they can lead to kernel panics, so one more case is no smp by qemu:arg and no <topology> line in the xml. Thanks Marking as minor, as the vm disks are not deleted, but some problems to rebuild the template if you have customized it and you didn't save it somewhere else. xml-template.txt
  5. ghost82
    Changed Status to Solved
  6. ghost82
    Fixed in 6.8.1-RC1, marking as solved.
  7. ghost82
    😳😳😳😳
  8. ghost82
    I was trying to share the "rootshare" parent folder, disabling single shares for appdata, domains, system and iso folders, so to list only the unassigned device and the rootshare folder. However, by following your indications, i.e. deleting part of the global section, disabling netbios, enabling shares for isos and domains and deleting the rootshare config, there is no more the described issue. Thanks
  9. ghost82
    Here it is, downloaded after steps 1-5 of above Bug report. Thanks
  10. ghost82
    Unraid version: 6.8.0 stable Bug replication: 1- Create a user, user1 2- Create a second user, user2 3- In unraid webGUI, Users-->user1, or Users-->user2, user access for domains and system folders is set to "No access" 4- In unraid webGUI, Users-->user1, change user access for domains and system folders to read/write and apply: settings are correctly saved and service restarted 5- In unraid webGUI, go back to Users 6- Enter user1 7- Change user access for domains and system folders to "no access" and apply: settings are correctly saved and service restarted 8- Go back to Users 9- Enter user2: settings for user access for domains and system folders is set to read/write instead of "no access" 10b- Change user access for user2 for domains and system folders to "no access" and apply: settings are correctly saved and service restarted 11b- Go back to Users 12b- Enter user1: settings for user access for domains and system folders is set to read/write instead of "no access" So the issue here is that I'm not able to reset user access permissions for domains and system folders to "no access" for both user1 and user2. Are you able to replicate? Not sure if it's minor or annoyance, since there is the following workaround: 1- Enter terminal of Unraid server 2- cd /boot/config/shares 3- nano domains.cfg 4- delete the user in shareWriteList so it is shareWriteList="" then save CTRL+O, Enter, and exit CTRL+x 5- nano system.cfg 6- delete the user in shareWriteList so it is shareWriteList="" then save CTRL+O, Enter, and exit CTRL+x Now if you go back into the unraid WebGUI both user access for domains and system folders for both users, user1 and user2 are back to "no access" Additional info: Everything is good if settings are applied from the Unraid Webgui in menu Shares-->domains-->SMB user access and Shares-->system-->SMB user access
  11. ghost82
    Hi all, strange thing happening on my unraid server: Unraid version: 6.8.0 stable Unassigned devices version: 2019.12.29a My setup is as following: Users: root, user1, user2 SMB settings: [global] access based share enum = yes min protocol = SMB2 client min protocol = SMB2 client max protocol = SMB3 signing_required = no include = /tmp/unassigned.devices/smb-settings.conf [rootshare] path = /mnt/user browsable = yes valid users = user1, user2 write list = user1, user2 In unassigned devices settings both users user1 and user2 have read/write permissions. and SMB security is enabled. The unassigned device disk is mounted and shared over SMB. The client is a mac os Catalina 10.15.2 VM inside unraid. Bug report: 1- Mount and share the unassigned device 2- Connect to the unraid server (tower) over SMB and login with user1 or user2: rootshare folder and the unassigned device are listed 3- Open the unassigned device folder: it's ok, user1 and user2 can browse, read and write files 4- Go back to tower, user1 or user2 is still logged in 5- Open the rootshare folder: folders/files on first level are listed, but user1 or user2 has no access permissions This does not happen if: 1- Mount and share the unassigned device 2- Connect to the unraid server (tower) over SMB and login with user1 or user2: rootshare folder and the unassigned device are listed 3- Open the rootshare folder: it's ok, user1 or user2 can browse, read and write files 4- Go back to tower, user1 or user2 is still logged in 5- Open the unassigned device folder: it's ok, user1 and user2 can browse, read and write files Are you able to replicate it or is it my problem? Thanks
  12. ghost82
    Correct, but only enterprise gpus, which cost a lot of money. Not possible with a p4000, sorry. Moreover, I'm quite sure nvidia grid is not compatible with the kvm hypervisor.
  13. ghost82
    @SpaceInvaderOne Several people are asking for support about graphical issues related to the os section of the xml template and since they're not reading the full thread 😫, which is quite short now, maybe it is better to add a notice in your first post related to the issue and solution?
  14. ghost82
    You should read the thread, the issue and its solution have been reported several time! Tip: start reading page 16. As far as your other question about "errors", the red line is ok, it means that you are running your vm with custom args, nothing wrong. About the yellow lines, it means that you are passing a custom arg for the cpu which is not supported by your cpu: more into details the unsupported feature is pcid, so you can remove it from your custom args, from: <qemu:arg value='Penryn,kvm=on,vendor=GenuineIntel,+invtsc,vmware-cpuid-freq=on,+pcid,+ssse3,+sse4.2,+popcnt,+avx,+aes,+xsave,+xsaveopt,check'/> To: <qemu:arg value='Penryn,kvm=on,vendor=GenuineIntel,+invtsc,vmware-cpuid-freq=on,+ssse3,+sse4.2,+popcnt,+avx,+aes,+xsave,+xsaveopt,check'/>
  15. ghost82
  16. ghost82
    Try reading and applying these, if you have not tried yet: However, for the navi a kernel patch should be required. Look at this thread too: Some people are working to make a patch for the navi reset bug, work is still in progress.
  17. ghost82
    Here there are some useful info:
  18. ghost82
    One is the clover efi partition, the other one is the mac os efi partition: never touch the mac os efi partition.
  19. ghost82
    There are no web drivers for mojave, nor catalina; this means you will not have cuda nor hardware encoding/decoding (h264/hevc), but metal will work (not for all, it will work for the following NVIDIA cards: GTX Titan (GK110), GTX Titan Black (GK110), GTX Titan Z, GTX 780/Ti, GTX 780, GTX 770, GTX 760/Ti, GT 740, GT 730, GT 720, GT 710, GTX 690, GTX 680, GTX 670, GTX 660/Ti, GTX 650/Ti, GTX 645 (Fermi), GT 640 (Kepler, GK107/208), GT 630 (Kepler, GK208), Quadro 410, Quadro K420, Quadro K600, Quadro K2000/D, Quadro K4000/D, Quadro K4200, Quadro K5000, Quadro K5200, Quadro K6000, Quadro NVS510). Note also that you can have instability/artifacts with GT 740, GTX 660, GTX 650/Ti and GTX 645 with the GK106 core.
  20. ghost82
    The one with the operating system installed You could boot into mac os with a working configuration, then use clover configurator and set in the boot section the following boot arguments: -v, keepsyms=1 and debug=0x100 After that power off the vm, make your changes to the "non working" configuration and boot your vm: you should see some verbose output which might be useful. Moreover, you can post your clover config.plist and xml template (AFTER REMOVING the osk value), and provide some more info about your configuration (passthrough, etc), kexts, etc.
  21. ghost82
    This is a third party app and you can use it to configure clover; you have a gui interface instead of manually editing the config.plist file. I can run my catalina vm with 16 cores, so 8 is not the maximum value, but my cpu is different. You should provide some clover logs to see where the problem is.
  22. ghost82
    My suggestion is to look at the details of the analysis, for example in latest 5.1.0 version geekbench is not able to pass some tests in my vm (I don't remember if it was for the cpu or the gpu, most probably the latter), and this is the cause of the lower benchmark score.
  23. ghost82
    You can try this: from windows, open a cmd with admin priviledges and run the make_bootable.bat file.
  24. ghost82
    It's not so different from Clover, both are based on a config.plist file for the settings. As I wrote start with the qcow2 image you can find in the thread I reported, make a backup of your clover qcow2 image, then replace that qcow2 image with that of opencore and try to boot. Finally you can use opencore configurator, an app with a gui to setup your settings, very similar to clover configurator. Start simple, maybe without passthrough, than add your components one at a time.
  25. ghost82
    Hi, I don't have the 5700xt so I can't test it directly. From what I read, it seems someone had luck with the 5700xt with opencore+lilu+whatevergreen (latest releases of course). A bug has been opened on github by the same "someone" who was not able to make the 5700xt to work with clover+lilu+whatevergreen but was able to get it to work with opencore. So your issue may depend on your clover configuration, or not, it may depend on clover itself. Remember that as of 1st september 2019 Acidanthera group is not testing its releases with Clover, but only with Opencore bootloader. So my advice is to switch to opencore and test your 5700xt. I also suggest to read both the following threads, where you will also find a base qcow2 opencore image to download: and the latest pages here: PS: maybe you need at least 10.15.1 for the drivers.
×
×
  • Create New...