kbnomad Posted March 28 Share Posted March 28 (edited) Hi, I've had my unraid for about 2 years now (I think), but I've been on ipvlan for the longest time, and switched back to macvlan when 6.11.5 came out and gave the workarounds; also, I couldn't get to my dockers' webgui with ipvlan, as well as WOL doesn't really work with it. I never really noticed random freezes ever since (at least not that I know of) but I haven't put my unraid server to sleep, until recently when I installed S3 Sleep again, since I have pfsense router fully set up already. I also changed my flash drive Last October 2023 to Samsung Bar Plus, but recently got faulty about a few days ago (got the "Unable to enumerate USB Device"); so I changed that. but for the last 2 days, I've gotten 3 random crashes/reboot/freezes. Initially I was thought it was the Dynamix S3 Schedule (inactivity) sleep that was crashing my system, but then I saw some call traces on my syslog. I already ram memtest again since 2 yrs ago, and no errors. Anyone know if this is the macvlan call trace issue or something else? I currently switched to ipvlan to monitor it , and let S3 Schedule to put the system to sleep and see if it crashes. I have my diagnostics uploaded for reference as well.. Thanks for any help. Quote Mar 27 17:51:22 kbnomadserver kernel: Call Trace: Mar 27 17:51:22 kbnomadserver kernel: <TASK> Mar 27 17:51:22 kbnomadserver kernel: \_\_schedule+0x5b2/0x612 Mar 27 17:51:22 kbnomadserver kernel: schedule+0x8e/0xcc Mar 27 17:51:22 kbnomadserver kernel: request\_wait\_answer+0xf7/0x1d3 Mar 27 17:51:22 kbnomadserver kernel: ? \_raw\_spin\_rq\_lock\_irqsave+0x20/0x20 Mar 27 17:51:22 kbnomadserver kernel: fuse\_simple\_request+0x1b0/0x1ef Mar 27 17:51:22 kbnomadserver kernel: fuse\_statfs+0xb0/0x132 Mar 27 17:51:22 kbnomadserver kernel: statfs\_by\_dentry+0x40/0x5e Mar 27 17:51:22 kbnomadserver kernel: vfs\_statfs+0x16/0x93 Mar 27 17:51:22 kbnomadserver kernel: user\_statfs+0x54/0x9f Mar 27 17:51:22 kbnomadserver kernel: \_\_do\_sys\_statfs+0x20/0x50 Mar 27 17:51:22 kbnomadserver kernel: do\_syscall\_64+0x6b/0x81 Mar 27 17:51:22 kbnomadserver kernel: entry\_SYSCALL\_64\_after\_hwframe+0x64/0xce Mar 27 17:51:22 kbnomadserver kernel: RIP: 0033:0x14e42aab6537 Mar 27 17:51:22 kbnomadserver kernel: RSP: 002b:00007ffd64f46fd8 EFLAGS: 00000202 ORIG\_RAX: 0000000000000089 Mar 27 17:51:22 kbnomadserver kernel: RAX: ffffffffffffffda RBX: 00007ffd64f47080 RCX: 000014e42aab6537 Mar 27 17:51:22 kbnomadserver kernel: RDX: 00007ffd64f483e8 RSI: 00007ffd64f46fe0 RDI: 000055814bf371f0 Mar 27 17:51:22 kbnomadserver kernel: RBP: 00007ffd64f46fe0 R08: 00007ffd64f483e8 R09: 000014e42ab9c3e0 Mar 27 17:51:22 kbnomadserver kernel: R10: 0000000000000000 R11: 0000000000000202 R12: 00007ffd64f483e0 Mar 27 17:51:22 kbnomadserver kernel: R13: 000055814bf391b0 R14: 00007ffd64f483e8 R15: 000055814bf00800 Mar 27 17:51:22 kbnomadserver kernel: </TASK> kbnomadserver-diagnostics-20240328-1148.zip Edited March 28 by kbnomad Quote Link to comment
JorgeB Posted March 28 Share Posted March 28 That doesn't look like the typical macvlan call trace, but retest with ipvlan, also enable the syslog server, to see if it catches anything else. Quote Link to comment
kbnomad Posted March 28 Author Share Posted March 28 (edited) 2 hours ago, JorgeB said: That doesn't look like the typical macvlan call trace, but retest with ipvlan, also enable the syslog server, to see if it catches anything else. Interesting. Thanks for the info. ipvlan doesn't have call traces on the syslog, does it? I mean does the call trace only show when there's an issue? my syslog server has always been enable way back when switch to ipvlan in 6.10. so you think the instabality might be something else? I really want the macvlan to work because I like to be able to access my dockers while I'm out of town (remote). Edited March 28 by kbnomad Quote Link to comment
JorgeB Posted March 28 Share Posted March 28 If you really need macvlan, you can still use, as long as bridging is disabled. Quote Link to comment
kbnomad Posted March 28 Author Share Posted March 28 31 minutes ago, JorgeB said: If you really need macvlan, you can still use, as long as bridging is disabled. Yes, Bridging was disabled when I used macvlan, and tried both bonding enabled/disabled. Just enabled bridging today as I have to switch to ipvlan for monitoring; so far, left the server inactive, and S3 triggered the sleep schedule, woke it up through pfsense no problem twice. Though, it might be too little of a time to conclude yet but I got 3 crashes in 2 days, and just so happened when I updated to 6.12.9, might be coincidence, IDK. Just to clarify, disabled Bridging is the setting under "Settings > Network Settings > Enable Bridging = No; Enable Bonding = Yes" , right? Because that's what I've been using for months, though, the Docker Delugevpn, is still set to bridge mode as that is the only option that works properly. Quote Link to comment
JorgeB Posted March 28 Share Posted March 28 8 minutes ago, kbnomad said: Yes, Bridging was disabled when I used macvlan Then the call traces would not be about macvlan. 8 minutes ago, kbnomad said: Just to clarify, disabled Bridging is the setting under "Settings > Network Settings > Enable Bridging = No; Enable Bonding = Yes" , right? Yes. Quote Link to comment
kbnomad Posted March 28 Author Share Posted March 28 1 minute ago, JorgeB said: Then the call traces would not be about macvlan. Yes. Thank you. So the instability isn't about macvlan. so Call traces that isn't macvlan is fine? I will try to switch back to macvlan and monitor the sleep activities. hopefully, it was just the faulty flash drive which I had to change. Quote Link to comment
kbnomad Posted March 28 Author Share Posted March 28 just switched back to macvlan, and the system froze again as soon as the S3 sleep plugin triggered for inactivity. prior to switching (macvlan), the S3 sleep plugin triggered for inactivity twice, and was able to WOL twice, no problem; even went to sleep while the macOS VM was on (not ideal) but system didn't freeze. Will try to use ipvlan a couple more days and see if it freezes or not, then switch back to macvlan. Also, experienced the crashes/freezes after updating to 6.12.9 Quote Link to comment
JorgeB Posted March 29 Share Posted March 29 13 hours ago, kbnomad said: so Call traces that isn't macvlan is fine? It depends, but generally any call trace means that there is a problem, hardware or software, though some can be more serious than others. Quote Link to comment
kbnomad Posted March 30 Author Share Posted March 30 On 3/29/2024 at 5:15 AM, JorgeB said: It depends, but generally any call trace means that there is a problem, hardware or software, though some can be more serious than others. Oddly enough, haven't had call traces since then; Also, didn't have crashing issues until today, but I suspect it's because of the S3 Sleep auto inactivity scheduling, even though I'm running win 11 VM and copying files. It doesn't crash when there's no VM running and it goes to sleep, but it tries to go to sleep even when the VM is running, prolly because I have the NVME Controller and/or SSD Drive passed through. So it might be the culprit. 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.