Ask Your Question

echiu's profile - activity

2016-05-14 15:10:18 -0600 received badge  Necromancer (source)
2015-07-29 09:17:28 -0600 answered a question os-ansible-deployment execution error

Most likely a library path issue. You can try updating your ANSIBLE_LIBRARY path to include /etc/openstack_deploy.

2015-07-29 09:07:54 -0600 answered a question Poor network performance for traffic that hit "Floating IP"

Some ideas to try:

  1. You can run traceroute on the compute node to the guest VM IPs. This will show the number of hops and the time per hop.
  2. Generate a report from your routes on network traffic

Is the network latency symptom intermittent or consistent?

2015-07-29 02:46:17 -0600 received badge  Nice Answer (source)
2015-07-28 17:18:51 -0600 answered a question How to list private IPs in use

You can login in as admin and use Horizon to list the private subnet used by each tenant network. Otherwise you would have to login to the Openstack MySQL database as root user and run database queries to list all tenant networks.

2015-07-28 17:15:49 -0600 answered a question Unable to complete operation for network. The IP address is in use

You are using Cloudify for orchestration instead of native Openstack tools. It is likely that Cloudify maintains its own database to track IP addresses, you might want to check their documentation to get details.

Try running nmap -sP xx.10.70.xx to confirm if the IP is in use.

2015-05-01 01:04:44 -0600 answered a question keystone design documents

To use Keystone APIs call from your web application, what you need is the API documentation. Unless you need to develop custom version of Keystone, you don't need the architecture documentation or build it from source on github.

Test our your API calls using curl, per examples on this doc: http://docs.openstack.org/developer/k...

Once you have successfully tested the API calls, you can embed them in your web application.

2015-04-03 12:52:34 -0600 answered a question Automate - creating snapshot of Instances and Volumes

You can use ansible playbook to execute openstack Nova CLI commands. For example, see this module for snapshots you can reference from your playbook: https://github.com/weshayutin/nova_sn...

2015-04-03 12:20:23 -0600 answered a question How can one implement ACLs in Neutron?

There is an old blueprint for Quantum ACL that was never approved: https://blueprints.launchpad.net/neut...

It seems the current direction is to have central ACL through keystone rather than have each project have its own ACL. You can try a bottoms up approach and implemented for your own Neutron driver and propose it as part of the specification for Neutron.

2015-04-01 00:52:36 -0600 commented answer Configure OpenStack on a WiFi-only system?

On a wifi connection, you don't have eth0, I think the installation scripts assume you have a static IP hence bridging with eth0 fails as the other commenter pointed out.

2015-04-01 00:52:13 -0600 commented answer Configure OpenStack on a WiFi-only system?

On a wifi connection, you don't have eth0, I think the installation scripts assume you have a static IP hence bridging with eth0 fails as the other commenter pointed out.

2015-03-25 01:55:32 -0600 received badge  Nice Answer (source)
2015-03-24 16:35:32 -0600 answered a question Configure OpenStack on a WiFi-only system?

There is a difference in the physical NIC configuration and the virtual NIC configuration. The virtual NICs are part of an network overlay that does not exist outside your node.

Please let the hypervisor and install tools manage the vNICs in a virtual installation and do not use the network device manager on your node. You will not be using the IP on your physical NIC.

See this documentation for a RDO virtual installation example: https://www.rdoproject.org/Deploying_...

2015-03-23 17:36:55 -0600 answered a question Does swift 1.0 supports object versioning ?

The API documentation is not complete, so you should verify this directly by checking in the configuration file.

Add object version /etc/swift/container-server/container-server.conf

To enable object versioning, the cloud provider sets the allow_versions option to TRUE in the container configuration file.

This feature has been in Swift since at least 2012, so there is a good chance it is in v1.

2015-03-23 17:01:50 -0600 answered a question DevStack image pull hanging during ./stack.sh

Try to see if you can download the images directly: wget http://mirrors.kernel.org/fedora-alt/...

It seem to be a network issue, there are many possible causes: 1. your ISP is throttling bandwidth 2. router configuration 3. network instability

2015-03-23 16:55:07 -0600 answered a question swift object migration from one devstack to another devstack

In devstack, the data for Swift is stored in the source tree by default (in $DEST/swift/data) and can be moved by setting SWIFT_DATA_DIR. The specified directory will be created if it does not exist.

To migrate your data, copy this directory from the old devstack instance to the new instance.

You can change this in the localrc configuration

SWIFT_DATA_DIR=$DEST/data/swift
2015-03-23 16:45:02 -0600 answered a question Glance API shows authentication errno 1 during uploading image to swift's container

It appears you may have issues with glance/swift integration in your configuration. The following configuration sample of /etc/glance/glance-api.conf is taken from a working setup for your reference.

## NB: Unpolished config file
## This config file was taken directly from the upstream repo, and tweaked just enough to work.
## It has not been audited to ensure that everything present is either Heat controlled or a mandatory as-is setting.
## Please submit patches for any setting that should be deleted or Heat-configurable.
##  https://git.openstack.org/cgit/openstack/tripleo-image-elements

[DEFAULT]

# Introducing logging config for JSON logging
log_config_append=/etc/glance/api-logging.conf

# New style hierarchic (glance.api) params
default_store = file

known_stores = glance.store.filesystem.Store, glance.store.swift.Store

bind_port = 9292

sql_connection = mysql://glance:c57a7c33394fef60c0de626946a054b8bf80bf6a@localhost/glance

registry_port = 9191

rabbit_use_ssl = false
rabbit_userid = guest
rabbit_password = guest
rabbit_virtual_host = /
rabbit_notification_exchange = glance
rabbit_notification_topic = notifications
rabbit_host = 127.0.0.1

filesystem_store_datadir = /mnt/state/var/lib/glance/images/

swift_store_auth_version = 2
swift_store_auth_address = http://127.0.0.1:5000/v2.0/
swift_store_endpoint_type = internalURL

swift_store_container = glance
swift_store_create_container_on_put = True
swift_store_large_object_size = 5120
swift_store_large_object_chunk_size = 200
swift_enable_snet = False

delayed_delete = False
scrub_time = 43200
scrubber_datadir = /mnt/state/var/lib/glance/scrubber

image_cache_dir = /mnt/state/var/lib/glance/image-cache/

[keystone_authtoken]
auth_host = 127.0.0.1
auth_port = 35357
auth_protocol=http
auth_uri = http://127.0.0.1:5000/v2.0
admin_tenant_name = service
admin_user = glance
admin_password = 4d73a373336934bd36ef1d0ccac41e616eccc0ba

[paste_deploy]
flavor = keystone
2015-03-23 16:30:14 -0600 answered a question HP Helion community virutal install fail

Re-run the create seed.

You may need to re-run the install script multiple times since bash scripts does not handle every error condition and some child processes may terminate causing the parent process to halt.

Also use nmap to scan your network to ensure that there are no pingable IPs in the range from 192.0.8.xxx to 192.0.15.xxx. This will cause undercloud/overcloud provisioning to fail.

I suggest you use a narrow IP range for overcloud then what is set in the default configuration at /root/tripleo/tripleo-incubator/scripts/hp_ced_setup_net.sh then copy over this file to the seed node, then start the provisioning for undercloud and overcloud.

For example, I use the following bash script on the seed:

#!/bin/bash

# edit config file with correct overcloud IP range in seed node at /root/tripleo/tripleo-incubator/scripts/hp_ced_setup_net.sh 

export UNDERCLOUD_NTP_SERVER=0.us.pool.ntp.org
export OVERCLOUD_NTP_SERVER=0.us.pool.ntp.org
bash -x /root/tripleo/tripleo-incubator/scripts/hp_ced_installer.sh
2015-03-23 16:17:53 -0600 answered a question Using OpenStack with IPv6 only?

Devstack installation scripts by default using IPv4 for IP assignments in DHCP. In order to change this you would have to modify the install scripts.

For IPv6 deployment on IceHouse, see this gist "Quick Guide to deploy OpenStack IceHouse on top of Ubuntu 14.04" https://gist.github.com/tmartinx/9177697

2015-03-23 16:08:02 -0600 answered a question migrate instances with Ubuntu 14.04 from Havana to Juno

This seems like it is related to the image itself, rather than the migration process. Please check the following in your Ubuntu 14.04 image: 1. are you booting from a volume? 2. if so, you may need to attach the volume in cinder before you boot the Nova instance.

If you can provide details from nova-conductor.log in /var/logs that would be helpful.

2014-12-13 02:14:26 -0600 received badge  Nice Answer (source)
2014-12-12 16:16:36 -0600 answered a question Why Fedora 21 cloud image is provided in raw format ?

If you look the download page for Fedora 21, click on the "Are you a Openstack user" link https://getfedora.org/en/cloud/download/

There is a qcow2 image for Openstack http://download.fedoraproject.org/pub...

2014-12-12 15:54:56 -0600 answered a question install ironic in ubuntu

According the source code for ironic.drivers.modules.ipmitool, the exception Missing 'ipmi_terminal_port' parameter in node's driver_info is raised when the port is not specified: http://docs.openstack.org/developer/i...

Specify the port in the driver info, and re-run ironic node-validate command.

2014-12-11 19:21:41 -0600 commented answer nova-network api "create network" error 404

If you are getting HTTP 404 error with the REST Console, then the web endpoint is invalid. You can find valid examples of Nova API calls in your log file at /var/log/nova/nova-api.log Use the CLI to invoke Openstack changes, then look in your nova-api.log for REST API examples.

2014-12-11 19:14:28 -0600 commented answer How to power on or start all VMs of HP Helion, after restart the server or workstation.

See Helion documentation on backup and restore for more information on the scripts: http://docs.hpcloud.com/helion/openst...

Which version of Helion are you running? It may be that these scripts are not available in older versions of Helion.

2014-12-11 19:07:47 -0600 received badge  Enthusiast
2014-12-04 10:45:48 -0600 received badge  Nice Answer (source)
2014-12-03 20:10:32 -0600 received badge  Editor (source)
2014-12-03 20:08:02 -0600 answered a question How to extend the time allowed for overcloud create

In tripleo/tripleo-incubator/scripts/hp_ced_overcloud.sh, you can adjust wait time in the following lines, depending on whether you are using ESX or not. Adjust the time parameters in the wait_for function, which is set to 30 minutes by default.

## #. _`Wait for Nova Compute`
##    ::
if [ "${OVERCLOUD_CLOUD_TYPE}" != "ESX" ] ; then
    wait_for 30 10 nova service-list --binary nova-compute 2\>/dev/null \| grep 'enabled.*\ up\ '
fi
### --end

## #. Wait for L2 Agent On Nova Compute
##    ::
# openvswitch agent will not run on ESX proxy compute. Check only on KVM compute node
if [ "${OVERCLOUD_CLOUD_TYPE}" != "ESX" ] ; then
    wait_for 30 10 neutron agent-list -f csv -c alive -c agent_type -c host \| grep "\":-).*Open vSwitch agent.*-novacompute\"" #nodocs
    ## wait_for 30 10 neutron agent-list -f csv -c alive -c agent_type -c host \| grep "\":-).*Open vSwitch agent.*overcloud-novacompute\""
fi
### --end
2014-12-03 19:58:04 -0600 answered a question How to power on or start all VMs of HP Helion, after restart the server or workstation.

Typically as an operator, you would use Horizon to control VMs in the overcloud. But if the seed and undercloud are down, then this is not an option and you have to restore the entire system state, or re-install Helion.

You can use the backup and restore scripts to restore system state, e.g.

# set environment variables..

cd /root/tripleo/tripleo-incubator/scripts/
./hp_ced_backup.sh --seed -f /root/backup
./hp_ced_backup.sh --undercloud -f /root/backup
./hp_ced_backup.sh --overcloud -f /root/backup

# restore seed
./hp_ced_restore.sh -S -f /root/backup/backup_seed -c /root/baremetal.csv

# restore undercloud
./hp_ced_restore.sh -U -f /root/backup/backup_undercloud -c /root/baremetal.csv

# restore overcloud
./hp_ced_restore.sh -O -f /root/backup/backup_overcloud
2014-12-03 16:06:26 -0600 answered a question nova-network api "create network" error 404

The Openstack CLI tools are python scripts which invoke REST APIs, so if you are able to create the network in CLI but not directly in your REST API call this tells us the following:

  1. The REST API endpoint exists.
  2. You are not creating the right URL in your REST API call, hence the HTTP 404 error.

First, try to validate the URL is correct and working in a developer tool such as REST Console https://chrome.google.com/webstore/de...

Then generate the URL and print in your Java code, but comment out the actual REST API call. Only when you are sure you are generating the correct URLs, then make the REST API call. It is easier to debug this way on client-side vs the server-side. To debug on the server-side, you would have to look at the openstack logs, but there are many logs in /var/logs and this is more difficult.

2014-12-03 15:57:33 -0600 commented answer Is there an heat orchestration template repo that is open and shared?

Thanks for the answer, but I'm looking for a community supported repo where contributors can provide common IaaS platform templates and shares with others.

2014-11-26 13:39:40 -0600 received badge  Teacher (source)
2014-11-25 17:08:10 -0600 received badge  Famous Question (source)
2014-11-25 17:08:10 -0600 received badge  Notable Question (source)
2014-11-25 13:40:38 -0600 answered a question unable to create new image

You can set property protection to control who has permissions to perform glance actions. The demo user may need to be granted permission to create new images.

For more details, see documentation on property protections http://docs.openstack.org/developer/g...

2014-11-25 13:31:48 -0600 answered a question How to define vCPU=2.0 GHz?

This setting is defined at the hypervisor level, typically in the configuration file for nova.

For example, using KVM the default hypervisor the Compute service enables you to control the guest CPU model that is exposed to KVM virtual machines.

In libvirt, the CPU is specified by providing a base CPU model name (which is a shorthand for a set of feature flags), a set of additional feature flags, and the topology (sockets/cores/threads). The libvirt KVM driver provides a number of standard CPU model names. These models are defined in the /usr/share/libvirt/cpu_map.xml file. Check this file to determine which models are supported by your local installation.

For more details, see http://docs.openstack.org/juno/config...

2014-11-21 00:18:35 -0600 received badge  Popular Question (source)
2014-11-20 13:57:47 -0600 asked a question Is there an heat orchestration template repo that is open and shared?

With common use cases in heat orchestration across different organizations, is there an open repo that everyone can use for sharing?

2014-11-19 22:12:08 -0600 answered a question Is it possible to run hlinux on VirtualBox ?

Yes, hLinux is a Debian-derivative OS and should run in Virtualbox. You may have issues if you try to install additional debian packages from public repos since it is designed to be as minimal as possible, and does not have many of the standard packages in Ubuntu.

2014-11-19 22:12:07 -0600 answered a question Can I install HP Helion distribution on Dell hardware?

Yes, it can work on Dell hardware, but you may need to modify the install process to reflect the server management tools on non-HP platform. The baremetal.csv is used by the shell script at install, and you may need to modify this shell for your platform if it is not on HP's supported list.