bjornatic Posted February 19, 2021 Share Posted February 19, 2021 Well. I just installed Wireshark and I'm playing with it. I recorded different file transfert from my Pi (100mbps) as from this source, the copy works fine even with a 9GB+ file. And then I recorded file transfert from my Unraid server. Small mp3 file : OK. 500meg .zip : OK. 5Gb ISO : started OK to my surprise, but it suddently crashed (and by crash I mean back to boot screen). Here are the last lines recorded : 16 beeing the unraid server, and 24 beeing my Big Sur VM. So I see nothing in praticular. I'm trying now to search the console for hints, but as I'm quite the beginner in this area, this may take a while ^^ Quote Link to comment
ghost82 Posted February 19, 2021 Share Posted February 19, 2021 (edited) 8 minutes ago, bjornatic said: Well. I just installed Wireshark and I'm playing with it. I recorded different file transfert from my Pi (100mbps) as from this source, the copy works fine even with a 9GB+ file. And then I recorded file transfert from my Unraid server. Small mp3 file : OK. 500meg .zip : OK. 5Gb ISO : started OK to my surprise, but it suddently crashed (and by crash I mean back to boot screen). Here are the last lines recorded : 16 beeing the unraid server, and 24 beeing my Big Sur VM. So I see nothing in praticular. I'm trying now to search the console for hints, but as I'm quite the beginner in this area, this may take a while ^^ Start logging before a file transfer and stop it when it hangs (hope you don't have a kp). Save the log and attach it somewhere. From what you are describing it seems an issue related to a cpu kp. Edited February 19, 2021 by ghost82 Quote Link to comment
bjornatic Posted February 19, 2021 Share Posted February 19, 2021 (edited) It does not hang anymore. It either passes or crashes. I do not get a transfert hang at 134.2meg followed by a error -8084 like I did on 11.2... I do not think a succesfull transfert log would be useful, right ? I was suspecting my samsung 980PRO nvme drive to be part of the problem as it only occures at speed (when transferring from my 100Mbps Pi, it all goes well). But copying from the unraid server to a USB drive via the VM leads to the same problem. I will tweak around my CPU settings (number and isolation) and see if there is any difference. Edited February 19, 2021 by bjornatic Quote Link to comment
ghost82 Posted February 19, 2021 Share Posted February 19, 2021 2 minutes ago, bjornatic said: It does not hang anymore. It either passes or crashes. I do not get a transfert hang at 134.2meg followed by a error -8084 like I did on 11.2... I do not think a succesfull transfert log would be useful, right ? I was suspecting my samsung 980PRO nvme drive to be part of the problem as it only occures at speed (when transferring from my 100Mbps Pi, it all goes well). But copying from the unraid server to a USB drive via the VM leads to the same problem. I will tweak around my CPU settings (number and isolation) and see if there is any difference. More difficult, sorry I didn't read well your message. If it hangs the log can be useful, but if it hangs with a kp...look at the kp window once restarted and try to understand where the problem could be.. I have also kp related to cpu but only at very high cpu loads: I assigned all cpu to the vm (and this should not be correct), so when the vm uses all cores to 100% sometimes it kp. Try to monitor also the cpu load when it fails. Quote Link to comment
bjornatic Posted February 19, 2021 Share Posted February 19, 2021 (edited) panic(cpu 3 caller 0xffffff800969a036): [kext.kalloc.512]: element modified after free (off:8, val:0x0000000000000001, sz:512, ptr:0xffffff9352a13400, prot:zero) 8: 0x0000000000000001 32: 0x0000000400000000 360: 0x0000000000000194 368: 0x0000000022477c09 376: 0x0000000000000194 384: 0x0000000022477c09 Backtrace (CPU 3), Frame : Return Address 0xffffffa07d552c40 : 0xffffff8008e8af3d [...] 0xffffffa07d553fa0 : 0xffffff8008e301f6 Kernel Extensions in backtrace: com.apple.filesystems.smbfs(3.5)[187ED778-E51A-34B6-81DE-9462540A2F8F]@0xffffff7fa2eaa000->0xffffff7fa2f1ffff dependency: com.apple.kec.corecrypto(11.1)[AE778828-78BF-365E-8C5A-2B5BCBE843DF]@0xffffff800c0a9000->0xffffff800c13afff dependency: com.apple.kext.triggers(1.0)[7A3DE42E-FDB4-3B20-A098-EE36FFE425E2]@0xffffff7fa2f25000->0xffffff7fa2f27fff This is the crash report... I read "CPU panic" and "com.apple.filesystems.smbfs"... My vm uses the 4 last cores (8 HT) and now I'll try with isolating them. Edited February 19, 2021 by bjornatic Quote Link to comment
ghost82 Posted February 19, 2021 Share Posted February 19, 2021 (edited) Ok, this seems the same type of kp I had from beta 7 big sur 10.16 (it didn't happen all the time, sometimes it hanged sometimes crashed with that kp <10%). I have no solution... Does switching to NFS is a no go for you?NFS seems to not have issues. Edited February 19, 2021 by ghost82 Quote Link to comment
bjornatic Posted February 19, 2021 Share Posted February 19, 2021 (edited) Does Time Machine backups work on NFS ? If yes, I could live without SMB... But I hate it not to have my VM fully fonctionnal without knowing why. 🙂 Anyhow, thank you for your help and concern. I really appreciate it. I'll report again if I find somthing new. Right now I'm creating a new fresh Big Sur VM and try to work on that. Edited February 19, 2021 by bjornatic Quote Link to comment
ghost82 Posted February 19, 2021 Share Posted February 19, 2021 (edited) 13 minutes ago, bjornatic said: Does Time Machine backups work on NFS ? It should, it's not officially supported by Apple but it should work. As a starting point try to read here (it's from 2016, but it should still be valid): http://m0agx.eu/2016/03/23/using-os-x-time-machine-with-nfs/ I confirm that the "fix" I posted does work for me, no more hangs or kp with file transfers, but smb is still unusable when running a vm from the mac os vm on a smb disc. <-- kp It fails when unpacking initramfs for a linux vm with the message "junk within compressed archive" and it kp. Running a vm from the vm on a nfs share works well. Edited February 19, 2021 by ghost82 1 Quote Link to comment
bjornatic Posted March 3, 2021 Share Posted March 3, 2021 So. I'm happy to report that I found a solution to my problem. For the record, using NTS shares worked a little better (no instant kp), but it evnetually lead to a kp. The solution was to switch the adaptor to from e1000-82545em to virtio-net. And now everything is smooth. I can do TimeMachine backups and also tranfert a 15GB file in 4min. I knew about vmxnet3 as an alternative to e1000, but I noticed that another VM of mine was using virtio-net.... So I gave it a try and it was such a simple solution that I was mad not having found it earlier. I still could not find relevant infos on the different possible network adaptet... vmxnet3, e1000-.., virtio-net... Does other exist ? Where could I find info about them ? 1 Quote Link to comment
ghost82 Posted March 3, 2021 Share Posted March 3, 2021 Thank you for sharing this! I confirm it's working fine for me too. With virtio-net I don't need anymore the line: socket options = TCP_NODELAY IPTOS_LOWDELAY SO_RCVBUF=65536 SO_SNDBUF=65536 in samba configuration, no kernel panics and files that I wasn't able to transfer are now ok. Moreover I'm able to run again my vm within the mac os vm through smb instead of nfs. So I'm changing right now the model of my interface for my internal network (vm <--> host) from e1000-82545em to virtio-net. I'm leaving the e1000-82545em for the 2nd interface with internet access bridge. Again thank you! 1 Quote Link to comment
ghost82 Posted March 3, 2021 Share Posted March 3, 2021 (edited) 2 hours ago, bjornatic said: I still could not find relevant infos on the different possible network adaptet... vmxnet3, e1000-.., virtio-net... Does other exist ? Where could I find info about them ? These are all the available devices: Network devices: name "e1000", bus PCI, alias "e1000-82540em", desc "Intel Gigabit Ethernet" name "e1000-82544gc", bus PCI, desc "Intel Gigabit Ethernet" name "e1000-82545em", bus PCI, desc "Intel Gigabit Ethernet" name "e1000e", bus PCI, desc "Intel 82574L GbE Controller" name "i82550", bus PCI, desc "Intel i82550 Ethernet" name "i82551", bus PCI, desc "Intel i82551 Ethernet" name "i82557a", bus PCI, desc "Intel i82557A Ethernet" name "i82557b", bus PCI, desc "Intel i82557B Ethernet" name "i82557c", bus PCI, desc "Intel i82557C Ethernet" name "i82558a", bus PCI, desc "Intel i82558A Ethernet" name "i82558b", bus PCI, desc "Intel i82558B Ethernet" name "i82559a", bus PCI, desc "Intel i82559A Ethernet" name "i82559b", bus PCI, desc "Intel i82559B Ethernet" name "i82559c", bus PCI, desc "Intel i82559C Ethernet" name "i82559er", bus PCI, desc "Intel i82559ER Ethernet" name "i82562", bus PCI, desc "Intel i82562 Ethernet" name "i82801", bus PCI, desc "Intel i82801 Ethernet" name "ne2k_isa", bus ISA name "ne2k_pci", bus PCI name "pcnet", bus PCI name "rocker", bus PCI, desc "Rocker Switch" name "rtl8139", bus PCI name "tulip", bus PCI name "usb-net", bus usb-bus name "virtio-net-device", bus virtio-bus name "virtio-net-pci", bus PCI, alias "virtio-net" name "virtio-net-pci-non-transitional", bus PCI name "virtio-net-pci-transitional", bus PCI name "vmxnet3", bus PCI, desc "VMWare Paravirtualized Ethernet v3" I think you can list them in a terminal with the command: qemu-system-x86_64 -device help or qemu -device help (most probably the second command will work in unraid) This will list all the available devices (not only network ones). Edited March 3, 2021 by ghost82 2 Quote Link to comment
bjornatic Posted March 25, 2021 Share Posted March 25, 2021 (edited) So, I just installed Big Sur 11.3 Beta 5 (I think... I lost track). Anyway its build 20E5217a. I wanted to try e1000-82545em again as virtio-net was working fine except when I put the system to sleep : on wake I had lost all network. And so far I do not see the kernel panic I had on previous betas. I haven't done extended testing yet, but so far so good. And I can put my system to sleep again and not lose network on wake. File transfert from the Unraid server file also seems faster. Edited March 25, 2021 by bjornatic 1 Quote Link to comment
ghost82 Posted March 26, 2021 Share Posted March 26, 2021 (edited) On 3/25/2021 at 10:20 AM, bjornatic said: when I put the system to sleep : on wake I had lost all network. I just tried with the stable release and yes, I confirm that if the vm goes into suspended state on wake all my networks are down, not only the virtio networks but also the network of my usb wifi dongle (passed through). In the energy settings I set for the test: Turn display off after: 1 min (just to try) Prevent computer from sleeping automatically when the display is off: Not checked Start up automatically after a power failure: Not checked Enable Power Nap: Checked After wake, in the system preferences all networks show with the "green dots", but there's no connectivity (for all). I'm not using sleep/wake on my vm, so it's ok for me, but to wake the vm I had to issue the command "virsh dompmwakeup VMname", from a second pc, or from the server with a kvm switch for the keyboard/mouse dongle, in my case pressing mouse or keys on the keyboard attached to the vm do not wake up the vm; out of curiosity, how do you wake up your vm? If with keyboard/mouse do you have an apple keyboard/mouse or third party? Edited March 26, 2021 by ghost82 Quote Link to comment
bjornatic Posted March 26, 2021 Share Posted March 26, 2021 I simply use the "Virtual Machine Wake On Lan" plugin. 🙂 1 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.