Hi,
using ACS with KVM and RBD/Ceph, it turns out that the user has chosed a
Compute Offering with volume type "Fat" provisioned. After he reinstall
the VM from the template using a Compute Offering with Volume "thin"
provisioned, the performance level was as high as the other VMs. Seems
like there is a huge performance difference between fat and thin,
however, I have not tested out the "sparse" option. I always thought
using RBD as primary storage there is no difference between fat and thin...
Regards,
Mevludin
Am 07.09.2022 um 15:01 schrieb Nux:
Hi,
Check that the templates were registered with the appropriate profiles
(ie of type CentOS7 etc), particularly if their storage and network
devices are Virtio ones.
You can do this with lspci on Linux - inside the VM - ethernet and
storage should present themselves as "virtio" or similar.
Another thing you can look at, is whether the qcow2 file they upload
is funky in any way, ie look at compat, cluster_size and compare with
a file in your cloud that you know works well. You can use "qemu-img
info file.qcow2" for this.
hth
---
Nux
www.nux.ro
On 2022-09-07 13:26, Mevludin Blazevic wrote:
Hi all,
some of our users have reported that after they uploaded QCOW2
templates to our ACS environment and started a VM from the template,
the VMs ran very slow. In contrast, VMs installed directly in ACS
using ISOs for example are very fast. I wonder if something in the
upload view was misconfigured by the users.
We are using KVM and in the upload view we can choose Root disk
controller, OS Type and other options like enabling HVM. Any ideas?
Best Regards
Mevludin
