Vr2Io

Never think Orico will mfg. such mid-end product, equip two AQC-113 NIC. Have nice starting.

I would say no matter VM or Docker are stable in Unraid, the keypoint are what VM or Docker you run. In long time ago, I run NVR (Windows VM) and Homeassistant VM in 7*24. there also several docker and non-always on VM. 8 core CPU pinned 4 for Unraid and docker, remain 4 for VM ( each VM allocate 3 and share those 4 core ). But currently I change NVR and HA in docker and change to a miniPC (6 core), no pinning and no VM. The original platform just run those non-always on VM, so it usually in power off. In general, you need tried, observe and plan youself. In each milestone, I keep little bit optimize, the last milestone were I build a multi-power source DC system for that miniPC ( with big battery capable to power up 10hrs+ ). I need it always on even in some longer power outage.

You can press ESC and got the pull down menu and perform different checking.

You can use some tools to check CPU / GPU power state. I use Corefreq for CPU always. Although the reading could be wrong for some CPU, but it still useful to know what power drawing from CPU, for exampe, we would found large power different when apply loading to CPU, it have 4 times differrent.

Exclude external DB, I also setting in similar way and solid running for 2yrs+

Strange, but I haven't biglittle architecture CPU .... It doesn't look like CPU relate. I highly recommend that, it use much less resources, restart docker also much simple and fast. But switch to docker, HA need start from scratch.

Interesting, a bad signal that memory system have problem, even such low memory clock rate. You can further test by another memory test utility to verify.

Pls further test if start array in maintenance mode then stop array, does this success ？ If still hang, does enclosure activity LED periodic flash ？

I think you should narrow down which one cause problem first ( or both ? )

That's fine, you could do it later. I agree if only three 12v wire for 24 disk doesn't enough ( in fact 5v also not enough ), module cable even make it worst. I use non-modual PSU and it have enough cable to connect backplane ( 4 independent molex for 16 disks ) If need I will modify i.e. soldering wire in PSU module backplane to fullfill my requirement. I always modify electric appliances. For your original question, NOROCO backplane need 5v provide, I don't found they have extra DC to DC converter on PCB.

Yes. Yes, too. I notice 10.55.83.10/24 use tag 4091, but you use br0, it should be br0.4091 ? But seem tag over 4000 are special !! Pls note, in Unraid eth0 / br0 means untag, and eth0.5 / br0.5 means tag. Does br0 will pass through to tag 4091 interface ? ( You should ignore above question as you said all valn can access WAN ) Below are my Unraid docker look like, so eth0 are untag and eth0.666 are tag 666 to router's LAN trunk port.

Simple as use buildin backup USB stick function, there a script also call that function to backup to Unraid specific path. As result, keeping periods version for rollback. I also keep latest backup in Windows PC, so very quick to recreate boot strick if need. I change to use miniPC as 7x24 Unraid for long time, only two M2 SSD use 20W ( no VM ). Other big rig will power on if I need.

For my understanding, Pihole`s eth0, actually is br0.5. Pihole wouldn't know what's setting outside docker. Pihole eth0 not means Unraid eth0.

So you have different issue actually, I use vlan router, so it rather simple to enable different VLAN to access WAN, i only control what local traffic cross VLAN for maximise security. I also setup Wireshark to shoot network problem, if no probe in different node, it is quite hard to troubleshooting, almost a dead loop if guess only.

Yes, br0 access br0.5 (pihole) shouldn't need enable host access. Then you should focus why console ( assume you said Pihole console ) why can't ping 10.55.5.1 gateway, they are in same subnet.

I also use one ethernet for all vlan, but I use macvlan and not enable bridge. You also need confirm Pihole GUI default accept other subnet to access, I use Pihole to serve different VLAN too, but I only route dns query between VLAN, I didn't family Pfsense as I use hardware router doing all routing.

Problem may be because you use ipvlan and due to one physical ethernet for all vlan, then host network access ( if access to br0.0 ) may need to enable.

Due to harddisk price raising and planning need, I buy two WD 20TB disk and shuck for use. The interesting was it should what model disk ? If according database, it is HC560 class. But after investigate, it should be HC590. The reason are The while label have wording "US7SAU260". The outlook not same as HC560 ( right pic ), even HC570 and HC580. SMART attributes have item "90" NAND mastrer, which since HC570 have that technology . I am happy for got the shuck disk with latest technology disk model, although the throughput even lower then normal HC550 disk ( less ~30MB ). HC590 are 11 platter model , then what actually inside for 20TB HC590 disk ?? Many bad head / platter ?? The only bad news is one of my disk insert to a bad connect cage and generate UDMA CRC error record, it break zero error in the main build disk collection.

Formatting data disk are on the fly, problem look like something struggling in background, pls try reboot and not start array and generate diagnostics.

You have one NIC and network config bond eth0-3, pls try delete network config and try again ( post new diagnostics if problem not solve ) rm /boot/config/network.cfg rm /boot/config/network-rules.cfg # Optional, but recommended Another issue which not relate network problem, does total SATA device are 4 ? In the log, there are 5th SATA device (?) continue generate link alarm Dec 15 10:30:24 ClosetNAS kernel: ata28: limiting SATA link speed to 1.5 Gbps Dec 15 10:30:27 ClosetNAS kernel: ata28: COMRESET failed (errno=-32) Dec 15 10:30:27 ClosetNAS kernel: ata28: reset failed (errno=-32), retrying in 8 secs Dec 15 10:30:37 ClosetNAS kernel: ata28: COMRESET failed (errno=-32) Dec 15 10:30:37 ClosetNAS kernel: ata28: reset failed (errno=-32), retrying in 8 secs Dec 15 10:30:47 ClosetNAS kernel: ata28: SATA link down (SStatus 1 SControl 310) Dec 15 10:30:47 ClosetNAS kernel: ata28: limiting SATA link speed to 1.5 Gbps

But not make senses if mult channel enable will affect iperf test result. But iperf test not always reflect actual situation.

RIP Ronald, when I new to Unraid, he lighting me always. Respect for him. If something tag Dynamix, I will first think are you.

I am a light user, only have ~12 docker. I store docker in \tmp, so no such pain to pull all if reboot. So far no problem in every update.😁

Same white banner for grey theme. For spindown problem, I face once, disk will spinup soon after spindown and UD show have throughtput forever in some fixed figure (counter no change). After restart array, problem gone.

In CA, previous docker, about 10ish docker will list per page, if you tick select all, then current page docker will pull again.