VM WS12R2 slow

[antoniormjr]

Hello everyone, today I performed some comparative tests with the Proxmox VE 3.4 vs Xen SuSE Linux SLES 11 SP3. I have a VM Windows Server 2012 R2 with 30Vcpu Ram and 100GB with the terminal server installed receiving more than 90 simultaneous accesses.
The VM after starting consumes all the memory and 100GB lock, activated the Balloon and resolved the issue.
When the VM that started the virtio network interface oscillates causing the loss of packets on the network, changed the network interface to intel e1000 in which solved the problem.
Even solving the above problems did not get great performance compared to the Suse Linux Xen. The VM was installed completely by Proxmox GUI, did not import at all.

Someone would have any idea? KVM holds as much cargo as described above?

 

[manu]

Hi Antonio

can you post the output of qm config <your_VM_ID> ?

 

[antoniormjr]

Hi Manu,

Follow the settings below. I plan on testing the 4.0 beta version of Proxmox VE because it has a more updated version of qemu- engine, what do you think ?

Config PVE
proxmox-ve-2.6.32: 3.3-147 (running kernel: 2.6.32-37-pve)
pve-manager: 3.4-1 (running version: 3.4-1/3f2d890e)
pve-kernel-2.6.32-37-pve: 2.6.32-147
lvm2: 2.02.98-pve4
clvm: 2.02.98-pve4
corosync-pve: 1.4.7-1
openais-pve: 1.1.4-3
libqb0: 0.11.1-2
redhat-cluster-pve: 3.2.0-2
resource-agents-pve: 3.9.2-4
fence-agents-pve: 4.0.10-2
pve-cluster: 3.0-16
qemu-server: 3.3-20
pve-firmware: 1.1-3
libpve-common-perl: 3.0-24
libpve-access-control: 3.0-16
libpve-storage-perl: 3.0-31
pve-libspice-server1: 0.12.4-3
vncterm: 1.1-8
vzctl: 4.0-1pve6
vzprocps: 2.0.11-2
vzquota: 3.1-2
pve-qemu-kvm: 2.1-12
ksm-control-daemon: 1.1-1
glusterfs-client: 3.5.2-1

Config VM
bootdisk: virtio0
cores: 30
memory: 102400
name: tsvdi
net0: virtio=A6:7F:8A:01:C9:A6,bridge=vmbr0
numa: 0
ostype: win8
smbios1: uuid=facd6ce7-7412-4c42-b472-f62eb459fd5d
sockets: 4
virtio0: vms_proxmox_iscsi_lvm:vm-133-disk-1,size=100G

 

[WaltervdSchee]

Hi Antonio,

Checking your config for the vm, it looks like you've configured 120 Vcpu's ?
AFAIK the number of cores is multiplied by the number of sockets you configure, this may perhaps be the cause of some scheduling issue ?

 

[antoniormjr]

Hello, I made the changes put four sockets with 10 VCPU.I realized the server start and after receiving more than 40 connections via Terminal Server Windows 2012 began to lose network packets and crashed.

 

[fortechitsolutions]

Hi, just curious,

- did you play with settings, for example, drop vCPU to 8 or 16 cores and see if any change of behaviour?
- did you drop ram, maybe to a more modest setting such as 48 or 64gb, and see if any change of behaviour ?
- did you stress test network traffic in any other means to see if this can also trigger problems? (for example, running large nic-saturating file transfers for sustained period of time, or running NIC benchmark tool to generate sustained NIC traffic)
- did you stress test local DiskIO to see if that can also simulate/trigger problems (ie, run a local disk bench tool and let it grind for a while, see if sustained high activity on disk also causes issues ?)

I've got one client site running 3.X latest proxmox, with a Server 2012r2 VM / acting as a general-purpose term server (albeit, I am using "ThinStuff" 3rd party term services toolkit, not the MS RDS/Termserver stuff). I believe their VM is running ~32-48gb ram and ~12-16cores (approx) and the office has ~25 concurrent users doing general line-of-business office productivity / inventory management apps. Runs like a tank and they never complain. Super stable. Just using regular VirtIO NIC and HDD drivers. But from sound of it this is quite a bit lighter deployment than the one you are working with. But it is a reference data point that might be of interest.

Tim

 

[udo]

Hi Manu,

Follow the settings below. I plan on testing the 4.0 beta version of Proxmox VE because it has a more updated version of qemu- engine, what do you think ?

Hi,
perhaps you should update your system first? 3.4 is actual - do you have the nosubscriptions-repro enabled?

Like Walter and Tim allready wrote about vcore - if your VM use more cores than the host has, your VM will be really really slow.

How many cores your system have? Please post the output of following command:

grep processor /proc/cpuinfo | wc -l

If you have more than one socket (on the node) - or an AMD-cpu with more than one CPU on one die - you should enable numa.

Udo

 

[manu]

Hi,

How many cores your system have? Please post the output of following command:

grep processor /proc/cpuinfo | wc -l

If you have more than one socket (on the node) - or an AMD-cpu with more than one CPU on one die - you should enable numa.

Hi Udo
Btw we can also use nproc for that, it think it is in core-utils since Debian 7 and following.

 

[udo]

Hi Udo
Btw we can also use nproc for that, it think it is in core-utils since Debian 7 and following.

Hi,
ahhh - learn something new every day...

Udo

 

[antoniormjr]

Hello everyone.

Problem Solved. Last night I performed the installation of Proxmox PVE 4.0 Beta, I performed the procedures of (apt-get update, upgrade and dist-upgrade), Bond set up the network interfaces and following the restore of VM2012. Today during morning finalized with network stress tests highly successful with 90 RDP connections on the server and very stable.

I recommend everyone to migrate your PVE.

The following configuration PVE

Proxmox-ve: 4.0-10 (running kernel: 4.2.0-1-pve)
pve-manager: 4.0-36 (running version: 4.0-36 / 9815097f)
pve-kernel-3.19.8-1-pve: 3.19.8-3
pve-kernel-4.2.0-1-pve: 4.2.0-10
lvm2: 2.02.116-pve1
corosync-pve: 2.3.4-2
libqb0: 0.17.1-3
pve-cluster: 4.0-17
qemu-server: 4.0-23
pve-firmware: 1.1-7
libpve-common-perl: 4.0-20
libpve-access-control: 4.0-8
libpve-storage-perl: 4.0-21
pve-libspice-server1: 0.12.5-1
vncterm: 1.2-1
qemu-kvm-pve: 2.4-5
pve-container: 0.9-18
pve-firewall: 2.0-11
pve-ha-manager: 1.0-5
ksm-control-daemon: 1.2-1
glusterfs-client: 3.5.2-2 + deb8u1
lxc-pve: 1.1.3-1
lxcfs: 0.9-pve2
cgmanager: 00:37-pve2
criu: 1.6.0-1
zfsutils: 0.6.4-pve3 ~ jessie

The Server 2012 configuration
bootdisk: virtio0
Color: 16
IDE2: none, media = cdrom
Memory: 51200
name: tsvdi
net0: virtio = FA: 4D: A3: DE: A9: 40, bridge = vmbr0
a: 0
ostype: win8
smbios1: uuid = 8723f33e-9a45-4149-a8d7-650cbb9e3ae6
sockets: 2
virtio0: Local: 101 / vm-101-disk-1.Raw, size = 100G

Manu,

Run following command on the server.
grep processor / proc / cpuinfo | wc -l
64