I migrated from core to scale and it has been a nightmare for me. Been using core since 2020 and has been smooth sailing. Figured out a few things. One other issue has just come back as well. Was so fed up, that I installed Scale from scratch and made all the setting changes.
I have started to add more network interfaces so that certain VLANS have direct access to my TrueNAS systems without having to go through the firewall. Had one big issue which was once I added a 3rd network connection, my ping times would go < 1ms to as high as 800ms. Seems TN did not like me having mixed network ports (on-board and PCIe card or some on a 1Gb card and some on the 2.5Gb card). Once I put everything on the 1Gb card pings are back to < 1ms.
Another issue is that once I assign more than two networks to the system(s) the following happens. Whether the system is idle or I am making changes, the screen show the little cube logo for a bit, then it shows a dialing box/window saying" Connecting to TrueNAS… Make sure the TrueNAS system is powered on and connected to the network. with a thin progress like bar at the top. That show for a few seconds. Then it says, “Logging in” and I am back at the screen I was at before it happens. It keeps on doing this every so often. Tried searching but found nothing. Not even sure how to describe it for a simple search term. All the hardware is the is the same. Doing continuous ping tests to see if no response coincides with what is happening.
So far, not seeing anything in the logs that match in time when it is occurring.
Running quality HW.
It would be helpful if you shared details on what this quality hardware actually is and how you it up, especially the network.
All we know is that you have the builtin NIC plus two PCIe NICs, possibly using the 1GbE one right now, some VLAN magic is happening, an unknown firewall exists, and that you are having some kind of connectivity issue.
Supermicro X11SPH-nCTF MB with IPMI and 1 x 10GbE NICs, 164GB of ECC RAM, Intel Xeon Bronze 3204 CPU. Quad 1Gb NIC card, Quad 2.5Gb NIC card. Supermicro SC386A-R920B chassis, Seagate EXOS, Ironwolf Pro HDDs. Yes, using the quad 1Gb card for all my current network connections. I have several VLANs to segment my device, trust, guest, IoT, Not, Storage, AV (Audio/Video) and a few more. I am using pfSense as my firewall and Cisco CBS managed network switches. The extra NIC cards were installed before I migrated from core to scale. I had to reconfig my rsync tasks as I was using the module methos with core and that is not available in scale, so using SSH methos and that is working. I have 3 active network connections to my TrueNAS main unit and 4 on my backup unit. My management VLAN, trusted VLAN, AV VLAN to both and also Storage VLAN to my backup unit. On the system that is my TN backup, also running a Supermicro board with ATOM Soc. Has IPMI and quad 1Gb NICs on board.
I did notice something. I have 3 connections to the GUI, each on a different IP according to the IP for the VLAN. Now for the computer that is on the trusted VLAN and the session that is connected to the trusted VLAN IP of truenas, the session is stable. If i switch to the browser tab for a session on one of the other subnets, the same screen cycle happens.
I have three system running Proxmox VE along with other system and don’t have any issues with connectivity or strange behavior like I am having with the TN migration.
