SimonF

Does it do it when you just run nvidia-smi -q -x as that is the command the GPUstat is running.

Click on name there is an option there.

Can you provide metrics when the GPU is in use? i.e. transcoding?

Open a terminal root@computenode:~# ls /tmp/gpudata* /tmp/gpudata0000:00:02.0 root@computenode:~# ls /tmp/gpujson /tmp/gpujson root@computenode:~#

Can you provide two files from /tmp gpujson and gpudatapciid that matches the a380

It needs a newer kernel which will be in 7.3. intel_gpu_top will not work as b60 uses xe driver and not i915. They are talking of adding support to top for xe.

No worries glad you got it fixed.

Other option wouldbe to create vlans on eth2. this will also create a brx.vlan you can use in the vm.

I dont know the networking product, but is eth2 connected via an access port or a trunk. It would like it is a trunk an because there is no native vlan set all untagged traffic is just going on to vlan 1 I assume which is where the main connection is. Are you able to get a list of the mac address and vlan assignment for the switch.

In the VM template did you change the nic to use the bridge connected to eth2?

7.3 has drivers built in for 8169. ethtool eth4 Settings for eth4: Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full 2500baseT/Full 5000baseT/Full Supported pause frame use: Symmetric Receive-only Supports auto-negotiation: Yes Supported FEC modes: Not reported Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full 2500baseT/Full 5000baseT/Full Advertised pause frame use: Symmetric Receive-only Advertised auto-negotiation: Yes Advertised FEC modes: Not reported Link partner advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Link partner advertised pause frame use: No Link partner advertised auto-negotiation: Yes Link partner advertised FEC modes: Not reported Speed: 1000Mb/s Duplex: Full Auto-negotiation: on master-slave cfg: preferred slave master-slave status: slave Port: Twisted Pair PHYAD: 0 Transceiver: internal MDI-X: Unknown Supports Wake-on: pumbg Wake-on: d Link detected: yes lspci -ks 83:00.0 83:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. Killer E5000 5GbE Controller (rev 04) Subsystem: Micro-Star International Co., Ltd. [MSI] Device 7e17 Kernel driver in use: r8169 Kernel modules: r8169

Does not look like they support IPMI https://github.com/henrygd/beszel/issues/657

Do you mean the dashboard. There is an option in settings for values to be shown on the dashboard.

So does the whole host crash or just the VM? if you are not getting anything in the logs or on the console it will be difficult to find what is causing the crash if it whole host.

Hi, there is no integration into SMB with this plugin. Snapshots shows on the tools page, there is no tab across the menu bar. Not sure what issues you mean with cron as you define when you want something to run. Please post anything on the support page. I have not activiely made any updates to this plugin as most people are now using ZFS. I would expect snapshots to be added to the product at some point in the future. Any suggestions for options to add support to SMB settings I will raise a feature request to see if it can be added as a toggle/option to enable shadow copy.

Need to have the info before the reboot as diags will not have the issues before the reboot. You need to setup a syslog server or look at the console wen it fails.

I cannot see anything jumping out in the Diags. do you have any errors show on the console or do you have an external syslog setup. It may be there is an issue with the VM reset the card causing the host to have an issue.

Is the 4 port card inuse by the host, maybe post diagnostics.

What type of restart do you mean. There is an option in the settings for restart, but it is either restart or stop due to the nature of how WOL works with a broadcast packet it can only toggle to between states. Each of the operations have a restart command you could use from the command line. LXC you have to stop and start, but docker and VMs have restart/reboot commands. i.e. root@computenode:~# virsh help reboot NAME reboot - reboot a domain SYNOPSIS reboot <domain> [<mode>] DESCRIPTION Run a reboot command in the target domain. OPTIONS [--domain] <string> domain name, id or uuid [--mode] <string> shutdown mode: acpi|agent|initctl|signal|paravirt root@computenode:~#

How have you configure them in USBM. Also provide diagnostics if you can.

In the terminal run sensors this is the location for most temps.

Support for Bxx cards will be in 7.3 along with ability to set sr-iov for nics and gpus. The plugin will still need to be used for i915 gpus as it still needs a modified driver.

You need to have 6.17/.18 kernels to support Bxx cards, Support is being added for sr-iov in 7.3.

Intel GPU top is for i915 driver Bxx uses the xe driver and not covered by top, there is talk about using for xe but no timescales.

Do you get any sense data for the new SAS drives? /usr/sbin/sdparm -C sense /dev/sdx replace x with the drive letters