[Qlustar-General] login-c and beo nodes do NOT start
I´m deploying a head-node in a Virtualbox VM with a FE VM and 2 demo nodes (beo-01/2). After follow the doc guide's instructions, i have the head node up with perfect state but login-c and beo nodes couldn't get started. Anybody knows the reason and how to fix that? Thanks.
"M" == mauro delossantos <mauro.delossantos@udc.es> writes:
Hi Mauro, M> I´m deploying a head-node in a Virtualbox VM with a FE VM and 2 M> demo nodes (beo-01/2). After follow the doc guide's instructions, M> i have the head node up with perfect state but login-c and beo M> nodes couldn't get started. Anybody knows the reason and how to M> fix that? Thanks. please note that Qlustar does not support Virtualbox. Especially running the FE and demo nodes VM will not work inside a Virtualbox machine, since they will try to start a KVM machine inside Virtualbox which will not work afaic. What you can try is to set up additional Virtualbox VMs that will boot via PXE from the Virtualbox Head. But I don't know how to do the Virtualbox networking properly, so that all VMs share a common network (the cluster network) and that PXE will work. Good luck, Roland
In a sense I encountered similar problems, although I am trying to install a Qlustar OS on several Ethernet-linked machines. With the recent Installer 11.0.1-0 the head and FE VM nodes started and I can connect to them via ssh and Qluman, although with the previous version login-c appeared to be dead and dns stopped working after a reboot. However, even now attempts to add compute nodes end with the following message: PXE-E51: No DHCP or proxyDHCP offers were received. dnsmasq.service is running, /var/lib/tftpboot exists but does not contain pxelinux.0 -> /usr/lib/syslinux/pxelinux.0 pxelinux.cfg -> /etc/qlustar/pxelinux.cfg and the only pxe* files in the / tree are /usr/lib/grub/i386-pc/pxe.mod /usr/lib/grub/i386-pc/pxechain.mod /usr/lib/grub/i386-pc/pxeboot.img /usr/lib/ipxe/qemu/pxe-ne2k_isa.rom /usr/lib/ipxe/qemu/pxe-ne2k_pci.rom /usr/lib/ipxe/qemu/pxe-vmxnet3.rom /usr/lib/ipxe/qemu/pxe-virtio.rom /usr/lib/ipxe/qemu/pxe-e1000.rom /usr/lib/ipxe/qemu/pxe-rtl8139.rom /usr/lib/ipxe/qemu/pxe-pcnet.rom /usr/lib/ipxe/qemu/pxe-eepro100.rom /boot/grub/i386-pc/pxe.mod /boot/grub/i386-pc/pxechain.mod What has gone wrong and should be done? Thanks
"A" == Alexey Yaremchuk <a.yaremch@yandex.ru> writes:
Hi Alexei, A> In a sense I encountered similar problems, although I am trying A> to install a Qlustar OS on several Ethernet-linked machines. With A> the recent Installer 11.0.1-0 the head and FE VM nodes started A> and I can connect to them via ssh and Qluman, although with the A> previous version login-c appeared to be dead and dns stopped A> working after a reboot. However, even now attempts to add compute A> nodes end with the following message: A> PXE-E51: No DHCP or proxyDHCP offers were received. this means that your compute node physically cannot contact the head-node. Maybe the wrong network interface is used on the compute node. You need to check your cabling. A> dnsmasq.service is running, /var/lib/tftpboot exists but does not A> contain A> pxelinux.0 -> /usr/lib/syslinux/pxelinux.0 pxelinux.cfg -> A> /etc/qlustar/pxelinux.cfg A> and the only pxe* files in the / tree are A> /usr/lib/grub/i386-pc/pxe.mod A> /usr/lib/grub/i386-pc/pxechain.mod A> /usr/lib/grub/i386-pc/pxeboot.img A> /usr/lib/ipxe/qemu/pxe-ne2k_isa.rom A> /usr/lib/ipxe/qemu/pxe-ne2k_pci.rom A> /usr/lib/ipxe/qemu/pxe-vmxnet3.rom A> /usr/lib/ipxe/qemu/pxe-virtio.rom A> /usr/lib/ipxe/qemu/pxe-e1000.rom A> /usr/lib/ipxe/qemu/pxe-rtl8139.rom A> /usr/lib/ipxe/qemu/pxe-pcnet.rom A> /usr/lib/ipxe/qemu/pxe-eepro100.rom /boot/grub/i386-pc/pxe.mod A> /boot/grub/i386-pc/pxechain.mod A> What has gone wrong and should be done? None of these files are needed, since we migrated to GRUB as the boot loader (with Qlustar 11.0.1 which you are using). Best, Roland
participants (3)
-
Alexey Yaremchuk -
mauro.delossantos＠udc.es -
Roland Fehrenbacher