HPE Proliant Servers

    • Official Post

    You don't need anything special. Any USB 3 or 3.1 board will do as long as it fits. These servers can take half and full height cards on X4 slots. Gorgeous server by the way, take care of it. Here's your spec documents:


    https://www.hpe.com/h20195/v2/GetHTML.aspx?docname=c04282582

  • For the bios update you will need the latest HP SPP. You can take it from this torrent:


    hash: a5e1c453f10d9fb95677a6c2e796fffe0ad6189c
    announcers: http://leechers-paradise.org, udp://tracker.opentrackr.org:1337, udp://tracker.coppersurfer.tk:6969, udp://tracker.leechers-paradise.org:6969, udp://zer0day.ch:1337, udp://explodie.org:6969


    Regarding the cooler i recommend using the upgraded HP version that's rated for a 65W TDP: 832667-001

  • Hello,


    I'm setting up a HP Gen 8 1610T server.


    I carefully followed instructions from this thread to boot from a ssd on the 5th sata port and it worked perfectly, thanks.


    I installed a complete debian and then OMV 3.0


    --
    I have a very newbie question about networking :


    In my final setup I would like to use the second rj45 port for my printer/scanner and give it access to the network.


    Currently every device is connected directly to the internet router and is working fine (network print and scan).


    My network would look like :


    internet router -----------(rj45 1) gen8 (rj45 2)------------ printer
    |
    |
    PC


    Obviously, the printer should be accessible from the PC (or any other device on my network)... and ideally from internet trough google print (wich I use sometimes from work).


    1. is it even possible ? I wonder why it wouldn't.
    2. how ? :)


    It would allow me to move both the server and the printer to a closet (space is limited in my appartment) with only one long rj45 wire.



    Currently ifconfig only shows eth0 and lo.
    /etc/network/interfaces is standard :

    Code
    # The loopback network interface
    auto lo
    iface lo inet loopback

    I don't know where to start since I never configured anything similar (just routers with dhcp, port forwarding and default stuff).


    Thanks in advance !

    • Official Post

    I have some pretty straight forward answers for you:


    1. Yes it is possible, but is very complicated to implement unless you know how
    2. You would have to use the Linux IP Tables to do routing from two interfaces


    Even if you could do it, is not actually practical and you add dependencies that you shouldn't have. Instead, get an inexpensive network switch and that way you can share your only RJ45 drop. It will be easier, simpler and quicker.

  • Hi qkzk,


    1. It is possible (but not configurable in the OMV-webGUI)
    2. You have to set up a bridge between eth0 and eth1. For that you have to edit /etc/network/interfaces like this:

    NIC-port 1 should be connected to your router, NIC-port 2 to your printer (that is reachable only, when your OMV-machine is running ;) )
    After a reboot it should work...
    ...if not, because your system cannot set up eth0 and/or eth1, you need a local console (monitor and keyboard or iLo-remote-console) to delete /etc/udev/rules.d/70-network.rules and give it another reboot.
    Btw: Midnight Commander (mc) is imho very helpful for such console-jobs...
    BR
    Jan

  • A technical question Spy Alelo:
    I run two Windows-7-Virtualbox-VMs on my ML10v2 under OMV3. The VMs with IPs in the same network as the host and iLo are set up to use a bridged network connected to br0, a bridge between eth0 (NIC-port 1) and eth1 (NIC-port 2).
    Why is iLo (on shared NIC-port 1 with own IP) not reachable for the VMs, but for every other machine in my network?
    BR
    Jan

    • Official Post

    I don't know why you have bridged the interfaces, but if they are both on the same switch you will have some major issues passing traffic for certain things and iLO would be one of them. You have to realize that the fact that iLO shares a port with the system's NIC doesn't mean it makes the port a switch. I have a feeling that the iLO ARP record may not be passed along or the wrong record is presented to the VMs.


    Again, I am not sure why you bridged them but I would troubleshoot this by removing br0 and only use eth0 (where iLO is by default). Remove eth1 from the network temporarily and re-assign eth0 as the interface to be used by the VMs.


    If the VMs can now access iLO, you may want to rethink your layout. Bonding the embedded interfaces to each other is not recommended since you will encounter yet again a similar issue with iLO due to potential ARP issues. Here are a few suggestions:


    1. If you want redundancy, add an additional NIC and create a bonded interface between the card and eth1. Use balance-tlb or balance-alb if you don't have a switch that allows you to create LACP groups. Do not assign eth0 to be used by your OS, but have it connected to the network anyway. This will let your switch handle all the ARP records and iLO should be reachable by anything, and yet you still have redundant interfaces.


    2. If you don't need redundancy, but still would like to have different networks to be manageable dynamically per each VM, then implement a couple of VLANs. This of course would require extra hardware such as a manageable layer2/3 switch and configuring the VLAN logical interfaces on eth0. iLO is VLAN happy as well, so you can have it anywhere you like.


    3. Intel NICs usually behave better than Broadcom when managing this type of stuff, but you could get any 2-port or 4-port NIC and disable the onboard one. You can bond the ports as I suggested previously from within the very same card, and iLO will STILL work on the 1st port of the embedded NIC even after disabling it. Basically it will act as a bridge for iLO only and every VM should still reach it.


    I have a feeling that you may have to explain better why you have created that bridge, but this is the best I could come up with the little information you gave me.

  • I don't know why you have bridged the interfaces, but if they are both on the same switch you will have some major issues passing traffic for certain things and iLO would be one of them. You have to realize that the fact that iLO shares a port with the system's NIC doesn't mean it makes the port a switch. I have a feeling that the iLO ARP record may not be passed along or the wrong record is presented to the VMs.
    ...I have a feeling that you may have to explain better why you have created that bridge, but this is the best I could come up with the little information you gave me.

    Because the question was not how you would connect devices, i gave you all necessary information - except the matter of course, that only NIC-port 1 is connected to the external switch. I have the feeling, that you should better just answer "i don't know" in case you don't know an answer.


    BTW: A bridge makes the host with the bridged NICs work as a switch - and another NIC or another external switch and also the additional consumed power are not free of cost...


    BR
    Jan

    • Official Post

    BTW: A bridge makes the host with the bridged NICs work as a switch. And another NIC or another external switch are not free of cost and as well the consumed power isn't.

    Just remember that he is trying to help you. Not everyone (especially those in the enterprise environment) think about saving a few dollars and a few watts. You can buy a energy efficient gigabit switch for $20 and it will consume 3 watts. That isn't a big cost and is a supported config for OMV unlike a bridge.

    omv 7.4.10-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.14 | compose 7.2.14 | k8s 7.3.1-1 | cputemp 7.0.2 | mergerfs 7.0.5 | scripts 7.0.9


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    • Official Post

    Because the question was not how you would connect devices, i gave you all necessary information - except the matter of course, that only NIC-port 1 is connected to the external switch. I have the feeling, that you should better just answer "i don't know" in case you don't know an answer.
    BTW: A bridge makes the host with the bridged NICs work as a switch - and another NIC or another external switch and also the additional consumed power are not free of cost...


    BR
    Jan

    Obviously you don't know how iLO works to prevent the issue, but I'll leave it at that since you seem to be so knowledgeable and capable of resolving it yourself. And thanks for the lesson on a useless bridge config. I have a feeling, that you should quit while you are ahead.


    If you ask a question I'll try to answer the best I can to help people in this forum. Knowing the purpose of what you are trying to do helps me and others to think of alternatives to your problem as well. Don't like the answer? Then look somewhere else, don't get clever with those trying to help out.

  • Not everyone (especially those in the enterprise environment) think about saving a few dollars and a few watts. You can buy a energy efficient gigabit switch for $20 and it will consume 3 watts. That isn't a big cost and is a supported config for OMV unlike a bridge.

    Especially those in the enterprise environment should think more about saving a few watts, because they waste power on my cost since private consumers pay much more for power to support the "poor" economy. Wasting power without thinking and without valuable benefit brought us nuclear power plants and unsolveable problems for tenthousands of years...

    Obviously you don't know how iLO works to prevent the issue, but I'll leave it at that since you seem to be so knowledgeable and capable of resolving it yourself. And thanks for the lesson on a useless bridge config.

    Except for the described little issue my "useless bridge config" works perfectly as expected, even when you negate its use. Because i don't know how and why iLO and/or its port sharing generates this issue i asked you as the HPE-expert. Obviously you also don't know it. That is no serious problem - i can live without an answer. How about you?


    Imagine you ask your mechanic, why your e.g. Chrysler doesn't run in fourth gear although it has five gears, and he answers that he doesn't understand, why you want to use the "useless" fifth gear instead of buying a Dodge with a much more comfortable automatic gearbox, that he would prefer. I have a feeling that you wouldn't sense that as helpful.
    I sensed your answers as unkind, arrogant and not helpful.


    BR
    Jan

    • Official Post

    Especially those in the enterprise environment should think more about saving a few watts, because they waste power on my cost since private consumers pay much more for power to support the "poor" economy. Wasting power without thinking and without valuable benefit brought us nuclear power plants and unsolveable problems for tenthousands of years...

    That is a nice dream but unrealistic in the enterprise world. Where the customer wants as fast as possible and/or has to be super safe, a few watts is not going to be considered. In your case, a 3 watt switch would be a better idea in my opinion. I would bet that most people trying to save single digit watts with their server are not using LED lighting in every part of their house/apartment/flat/whatever.


    I sensed your answers as unkind and arrogant.

    In his defense, your responses aren't any better.

    omv 7.4.10-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.14 | compose 7.2.14 | k8s 7.3.1-1 | cputemp 7.0.2 | mergerfs 7.0.5 | scripts 7.0.9


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    • Official Post

    Except for the described little issue my "useless bridge config" works perfectly as expected, even when you negate its use. Because i don't know how and why iLO and/or its port sharing generates this issue, i asked you as the HPE-expert. Obviously you also don't know it. That is no serious problem - i can live without an answer. How about you?

    Except it doesn't work as expected for you. You can make any assumptions you like about me, is not going to change or improve your situation. I actually find it hilarious that you think I don't know what's up lol.


    Go get yourself a real switch and stop trying to prove something you are not. You shouldn't be using enterprise equipment if you can't make heads or tails from the info I gave you. If you want to save 3 watts of power on a switch, you are doing it wrong. You are contradicting yourself by using an enterprise solution that has a monitoring SoC that is powered 24x7.


    Now let's stop this right here and move on to do something productive. There are more people that will truly appreciate any help they can get and you are on the way.

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!