Search results

it seems that for newer kernels LRO is set automatically by driver. But my problem reappeared in 2.6.32-4-pve. ethtool -k eth0 shows well a LRO off. So what? I found a duplicate here: http://forum.proxmox.com/threads/4923-22K-s-Speed-in-VZ-on-GB-Link So i consider this topic close. Thx, dietmar.

hi, i confirm the diag of garyb, i have the same issue :slow bandwidth into VM, fast in the NOD. Disabling rx-checksumming works, after re-enabling it super-speed is still there. I run pve 1.6 on ProLiant BL490c G6 with Broadcom Corporation NetXtreme II BCM57711E...

Be careful this workaround works with bnx2x version 1.52.1 but it doest'n work if you upgrade to 2.6.32-4-pve, the bnx2x module upgrades to version 1.52.15 and the disable_tpa param doesn't exists anymore :-/ so, still searching for a solution...

Everything's fine. Thank you for your excellent work.

Thanks for your efficient and fast anwser. Regards.

Hi all, i misunderstand some pveversion outputs: proxmox2:~# pveversion -v pve-manager: 1.5-10 (pve-manager/1.5/4822) running kernel: 2.6.24-10-pve pve-kernel-2.6.18-2-pve: 2.6.18-5 qemu-server: 1.1-16 pve-firmware: 1.0-5 libpve-storage-perl: 1.0-13 vncterm: 0.9-2 vzctl: 3.0.23-1pve11 vzdump...

hi, i'm trying to update from 1.5 to 1.6 but when i do atp-get update i get this warning: W: Impossible de récupérer http://download.proxmox.com/debian/dists/lenny/Release Unable to find expected entry pve/binary-amd64/Packages in Meta-index file (malformed Release file?) here is my...

i assume you are running a windows guest kvm vm! a the moment my tests showed poor performances with virtIO nic. Try e1000 driver. http://pve.proxmox.com/wiki/Paravirtualized_Network_Drivers_for_Windows

yes i read a lot about clustered filesystems but i have never done that before and i don't think PVE supports it for OpenVz. I prefer waiting for their HA solution, thinking at the actual work, it will probably be really good.

I did it! just for fun and it worked, don't know how but it did! i created a vdisk (vzStore) on my FC san, the i presented it to the 2 proxmox server (in cluster mode). On the master, i created pv, vg and then a lv (vzLv1). a formated as ext3 and mounted it on /var/lib/vz on master. of course...

i found the problem, the TPA is enbable by befaut in the bnx2x driver. "TPA unfortunately is a form of LRO. Unlike GRO, it is fundamentally incompatible with bridging. So it has to be disabled when bridging is in use. "so i created a file bnx-options in /etc/modprobe.d with options bnx2x...

ok, next investigation. It's not a switch issue, i installled PVE on another server model and it works like a charm. So i think it's a issue with the nic model or a driver support. the problem happens on a ProLiant BL490c G6 with 10Gb nic NC532i, but i works on ProLiant BL460c G1 with 1Gb nic...

thanks for this very good tutorial :)

Hi, sorry if this post is a duplicate but but fore sure i've been searching for 3 day without results. I am very impressed by your work and everything could be fine but the bascis doesn't work :-( my problem affects either OpenVZ or KVM in bidge mode (vmbr0),pve kernel 2.6.18 i also used kernel...