eckdd's profile - activity

Server 2016. Install fails after timing out connected via SSH upon rebooting after install the LIS. Meanwhile, the console hangs at the following:

https://goo.gl/photos/LGLK7f1YG6aG92RP6

v-magine.log: http://paste.openstack.org/raw/601688/ v-magine-console.log: http://paste.openstack.org/raw/601689/

I ran into a similar error. Fixed it by ensuring that the external IP address on the host was static.

I noticed that the console available through the OpenStack dashboard imbeds content hosted on the FreeRDP Web-connect server that is installed on the Hyper-V host. This can be problematic if you want to console into a VM through the OpenStack dashboard, but the controller running Horizon is placed behind a port forwarding NAT device, or in the case where you would not want the Hyper-V host to be directly accessible to clients accessing Horizon.

I was wondering if there could be a way to move the FreeRDP web access off of the host and onto the controller, so that only a single IP address need be accessible by clients to access the VM console.

I've attempted similar things before with other HTML5/RDP solutions, but it seems that the standard FreeRDP client may not support the passing of the VMid BLOB attribute. However, I recently ran across a Veeam project that seems to do exactly that.

Upon launching an instance through OpenStack on a Hyper-V host, the VM is name "instance-00000xx"--is there a way to have the prefix "instance" replaced with something else on a per-launch basis, particularly using the metadata when launching the instance? For example, if I launch 5 instances of a special Windows 7 build, I would like the VM names in Hyper-V to be similar to "Special-00000xx"

Also, is there a way to configure VM attributes like Hyper-V integration services or CD drive options at the launch of an instance?

Yeah, I did console straight into the box as soon as it failed. Didn't delete anything and also didn't even close the installer.

Don't think it executed completely, but here's what I got: http://paste.openstack.org/show/584272/ I had to install paramiko. /etc/nova/ does not exist. Also, I used 172.21.x.x (my internet-facing private network) as the External IP.

Also dropped the core count down to 4 right before the VM powered on and it failed in at the same place with same error.

I did not notice any differences between the 2012 R2 log and the 2016 log; both fail in the same place and give the same error. Ran the install as both the local admin and separate account with admin rights. VM has 6144 MB with 16 cores assigned, host has 128 GB, dual 8 core Opteron 6212's.

Ran this on the latest release of Server 2016, then installed 2012 R2 and got the same thing. Looks like it bombs out after the following:

 2016-09-30 11:18:52,552 DEBUG logging debug 1630 4728 AsyncLoopThread OpenStack config: {'oslo_messaging_rabbit': {'rabbit_hosts': '', 'rabbit_userid': '', 'rabbit_password': ''}, 'glance': {'api_servers': ''}, 'keystone_authtoken': {'admin_tenant_name': '', 'admin_password': '', 'admin_user': ''}, 'neutron': {'username': '', 'project_name': '', 'url': '', 'user_domain_name': '', 'auth_url': '', 'password': '', 'project_domain_name': ''}}

2016-09-30 11:19:24,816 ERROR logging error 1596 4728 AsyncLoopThread need more than 1 value to unpack

Here's the rest of the log: http://pastebin.com/L9ZZuxj0 console log: http://pastebin.com/CM5Aq4Gp Prior to this, CentOS rebooted after installing the Linux Integration Services. It fails after the exact same event in both 2012R2 and 2016.