If I want to add multiple network adapters (DHCP - Ethernet), the UCS always hangs on the second adapter. Example: Quad-port network card first port can be added immediately and without problems, second port UCS hangs with the message “Setting up IPvX network devices”.
Already tried several external (different) cards, the cards and NICs were all correctly detected and displayed during installation. When configuring the second NIC of a quadport card, however, it unfortunately never goes further.
Where can I find the UCS logs for the network settings I would have made?
we prefer English language here, so I will reply in English. A short translation first of your issue:
If I want to add multiple network adapters (DHCP - Ethernet), the UCS always hangs on the second adapter. Example: Quad-port network card first port can be added immediately and without problems, second port UCS hangs with the message “Setting up IPvX network devices”.
Already tried several external (different) cards, the cards and NICs were all correctly detected and displayed during installation. When configuring the second NIC of a quadport card, however, it unfortunately never goes further.
Where can I find the UCS logs for the network settings I would have made?
Before checking if this is an issue of UCS I would try to configure the interfaces first manually.
Check the naming for your interfaces by
root@backup:/proc/net# networkctl
WARNING: systemd-networkd is not running, output will be incomplete.
IDX LINK TYPE OPERATIONAL SETUP
1 lo loopback n/a unmanaged
2 ens192 ether n/a unmanaged
3 docker0 ether n/a unmanaged
3 links listed.
Now you can configure your additional interfaces manually ie with ifconfig: ifconfig ens193 192.168.192.42 netmask 255.255.255.0 up
Once it is up you should be able to ping any host in this network which indicates the network card itself is running fine.
Once confirmed it is running fine you might configure your additional interfaces through ucs-variables as follows:
ucr set interfaces/eth1/address: 192.168.9.10
ucr set interfaces/eth1/broadcast: 192.168.9.255
ucr set interfaces/eth1/ipv6/acceptRA: false
ucr set interfaces/eth1/netmask: 255.255.255.0
ucr set interfaces/eth1/network: 192.168.9.0
ucr set interfaces/eth1/start: true
ucr set interfaces/eth1/type: static
Check documentation if you prefer to set one of the interfaces to use a DHCP-assigned address.
Webinterface is stuck sometimes, any way to reset it? Tried service apache2 restart, but it did not work. Do I have to use any specific UCS-command for cmd to do this?
both are two completely different things which are confusing for anyone reading here later. So I’d prefer to create a new thread for these issues.
To do it on command line you should better use PuTTy to connect to your server instead of using the console directly (BTW: virtual machine [which hypervisor?] or physical?)
unfortunately I have not seen evidence about the cards being configured correctly by hand. If you have configured them by hand does everything work properly?
Additionally with what settings are you trying to configure your cards through UMC?
What are identical parts from the server you are trying to the i350 T4?
Instead of rebooting you can just “kill” the process by using it’s process ID (pid) (ie kill 6029 or kill -9 6029).
Hi, just tried with ucr set command. It worked with this commands:
ucr set interfaces/eth1/address='192.168.9.10'
ucr set interfaces/eth1/broadcast='192.168.9.255'
ucr set interfaces/eth1/netmask='255.255.255.0'
ucr set interfaces/eth1/start='true'
ucr set interfaces/eth1/type='static'
Its important to put the values in ’ ’ and to use = otherwise they wont work. If I configure them as mentioned above everything works, no 100% process.
Additionally with what settings are you trying to configure your cards through UMC?
Just selecting interfaces, next, dhcp selected, next finish.
What are identical parts from the server you are trying to the i350 T4?
I also noticed that despite the problem with the network card, the system is really slow, while debian 9.4 is running smoothly. If I use dmidecode on ucs console the lines are printed line by line, because of performance. How can I check the debian version, UCS 4.3.2 is using? It must be something thats caused by the ucs image.
and you have configured all of your six (or 5?) network cards like this, yes? By ucr commands? And it is working flawlessly? The commands you posted shows static configuration while you mentioned for UMC using dhcp. There is some mismatch…
Well, at least we have a workaround then as it appears you are able to configure your additional NICs, right?
But you need to provide further details about your network envrionment. I need to know the network addresses, netmasks, gateways for VERY SINGLE network card you are trying to configure.
From which network do you access the UMC? What is the network configuration of your first NIC?
I would guess it is some sort of routing issue possibly? Unsure, though.
So i figured out the problem of the console (directly via monitor) performance. Thats a problem with video-output. I resolved it with the following boot parameters:
video=vesafb:ywrap,mtrr:3
vesafb use the VESA Frame Buffer
ywrap enable display panning using the VESA protected mode interface
mtrr:3 use Memory Type Range Registers in write-combining mode
The network problem (configurations hangs…) was caused by configuring IP addresses of the same subnet on multiple interfaces. CPU still goes 100% but configuration is done successfully after a few seconds, CPU goes back to normal then.