Kunal Patil's profile - activity

2014-03-18 14:25:41 -0600 received badge  Famous Question (source)
2014-03-10 13:03:11 -0600 received badge  Notable Question (source)
2014-03-07 09:05:02 -0600 received badge  Popular Question (source)
2014-03-07 07:37:24 -0600 answered a question Cannot ping instance created in 2 node Openstack installation with neutron gre tunnelling

I do not know if I can answer my own question, but I have made some progress. Instead of using a third nic eth2, I have moved the eth0 configuration to br-ex.

CONTROLLER:

vi /etc/sysconfig/network-scripts/ifcfg-br-ex

DEVICE=br-ex

IPADDR=10.10.155.11

NETMASK=255.255.255.0

GATEWAY=10.10.155.254

ONBOOT=yes

vi /etc/sysconfig/network-scripts/ifcfg-eth0

DEVICE=eth0

ONBOOT=yes

HWADDR=38:EA:A7:90:0C:5C

ovs-vsctl add-port br-ex eth0 ; service network restart

Now I can provision instances with private and floating IP. I can access the instances using the floating IPs from the controller and the compute nodes.

I still cannot access the private IPs from the controller and compute nodes, or access the floating IPs from external network (The floating IP subnet is part of existing network, and the External Gateway is set accordingly). Please advice.

EDIT:

Fixed the issue. I was trying to create a floating IP with subnet 10.10.155.0/25 and using a smaller floating IP range. According to the suggestion at http://openstack.redhat.com/forum/dis... , "Your floating IP network should be defined with the same mask even if you are starting and ending the pool as a subset". After creating External network with 10.10.155.0/24, I am able to reach the floating ip instances from my network. Also, after adding the rule:

"route add -net 192.168.32.0 netmask 255.255.255.0 gw 10.10.155.31"

on my controller and compute nodes, I am able to login to the instances using the private IPs.

2014-03-07 04:13:28 -0600 received badge  Commentator
2014-03-07 04:13:28 -0600 commented question Cannot ping instance created in 2 node Openstack installation with neutron gre tunnelling

I saw "Failed to start LSB: Bring up/down networking". Network is unreachable (while trying to reach metadata URL).

2014-03-07 04:01:44 -0600 commented question Cannot ping instance created in 2 node Openstack installation with neutron gre tunnelling

My bad. I do not see a private IP assigned to the instance. Please see update question with the screenshot.

2014-03-07 03:50:32 -0600 commented question Cannot ping instance created in 2 node Openstack installation with neutron gre tunnelling

Yes on both counts. Internal IP for Instance is 192.168.32.2, while floating is 10.10.155.32. Router internal interface IP is 192.168.32.1, while external gateway is 10.10.155.31.

2014-03-07 01:36:13 -0600 received badge  Scholar (source)
2014-03-07 01:35:58 -0600 received badge  Famous Question (source)
2014-03-07 01:34:57 -0600 commented question 2 node Xenserver Openstack installation using RDO with Neutron OVS GRE

After a long fight, we gave up on trying to get openstack to work on xenserver. We are going to stick with KVM for now (had problems with Cinder and Neutron to get it working, not to mention the restriction with RAW images).

2014-03-07 01:33:06 -0600 received badge  Famous Question (source)
2014-03-07 01:28:29 -0600 asked a question Cannot ping instance created in 2 node Openstack installation with neutron gre tunnelling

ping 10.10.155.32 [floating IP]

PING 10.10.155.32 (10.10.155.32) 56(84) bytes of data.

From 10.10.155.11 [Controller node IP] icmp_seq=2 Destination Host Unreachable

I am trying to ping the instances created in a 2 node Openstack installation with neutron gre tunnelling. I am getting the following error: "Destination Host Unreachable"

Steps I took for installation and configuration are as follows. Please let me know what am I missing:

###################################################################################################################### Configure NETWORK

KVMController1 IPs: eth0: 10.10.155.11 eth1: 10.10.155.12 eth2: 10.10.155.13

KVMCompute1 IPs: eth0: 10.10.155.14 eth1: 10.10.155.15

Disable SELINUX

yum -y update

yum install -y http://rdo.fedorapeople.org/rdo-relea...

yum install -y openstack-packstack

packstack --gen-answer-file my_answers.txt

vi my_answers.txt [Make changes for GRE configuration]

packstack --answer-file=my_answers.txt

reboot

######################################################################################################################

Update the external bridge configuration [I tried using the OVSBridge and OVSPort options, but that resulted in all eth0, eth1 and eth2 IPs being inaccessible]

CONTROLLER:

vi /etc/sysconfig/network-scripts/ifcfg-br-ex

DEVICE=br-ex

TYPE=Bridge

ONBOOT=yes

BOOTPROTO=none

IPADDR=10.10.155.13

NETMASK=255.255.255.0

GATEWAY=10.10.155.254

IPV6INIT=no

NM_CONTROLLED=no

vi /etc/sysconfig/network-scripts/ifcfg-eth2

ONBOOT=yes

BOOTPROTO=none

PROMISC=yes

DEVICE=eth2

HWADDR=38:EA:A7:91:15:10

IPV6INIT=no

NM_CONTROLLED=no

ovs-vsctl del-br br-ex

ovs-vsctl add-br br-ex

ovs-vsctl add-port br-ex eth2

######################################################################################################################

http://docs.openstack.org/havana/inst...

neutron net-create Public1 --router:external=True --provider:network_type gre --provider:segmentation_id 2

neutron subnet-create Public1 --allocation-pool start=10.10.155.31,end=10.10.155.120 --gateway=10.10.155.254 --enable_dhcp=False 10.10.155.0/24

keystone tenant-list [To get my tenant ID I have created earlier]

neutron router-create Router1 --tenant-id 442c1ffbca794753aa0323923d937f98

neutron router-gateway-set 2a007868-54dd-462a-b425-cd912b2ab256 c59906b1-9157-483c-a7fe-33b742798397

neutron net-create --tenant-id 442c1ffbca794753aa0323923d937f98 Net1 --provider:network_type gre --provider:segmentation_id 3

neutron subnet-create --tenant-id 442c1ffbca794753aa0323923d937f98 Net1 192.168.32.0/24 --gateway 192.168.32.1

neutron router-interface-add 2a007868-54dd-462a-b425-cd912b2ab256 c5207cc1-3351-4437-be83-6bc0587d50bd

######################################################################################################################

Security Group Rules

Direction Ether Type IP Protocol Port Range Remote

Ingress IPv4 TCP 22 (SSH) 0.0.0.0/0 (CIDR)

Ingress IPv4 ICMP - 0.0.0.0/0 (CIDR)

######################################################################################################################

Created Fedora 19 instance from QCOW2 image, and added a floating IP.

######################################################################################################################

In case this would be needed (maybe I am missing Iface br-ex entry in the table?):

route -n

Kernel IP routing table

Destination Gateway Genmask Flags Metric Ref Use Iface

10.10.155.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0

10.10.155.0 0.0.0.0 255.255.255.0 U 0 0 0 eth1

169.254.0.0 0.0.0.0 255.255.0.0 U 1002 0 0 eth0

169.254.0.0 0.0.0.0 255.255.0.0 U 1003 0 0 eth1

169.254.0.0 0.0.0.0 255.255.0.0 U 1004 0 0 eth2

0.0.0.0 10.10.155.254 0.0.0.0 ... (more)

2014-02-11 23:32:33 -0600 received badge  Popular Question (source)
2014-02-11 23:32:33 -0600 received badge  Notable Question (source)
2014-01-15 05:33:10 -0600 asked a question Is Neutron OVS with GRE supported on Xenserver 6.2?

Just a question, as I am facing problems getting it to work. If it is, what is the configuration needed?

2014-01-15 05:31:16 -0600 commented question 2 node Xenserver Openstack installation using RDO with Neutron OVS GRE

I have progressed till the point where the VM is being generated, and I can see its console on Xencenter. But I cannot reach it through the private or the public IP. I believe that my setup is not correct. When I try to use XenAPIOpenVswitchDriver, I get a cannot import module error.

2014-01-15 05:28:58 -0600 received badge  Notable Question (source)
2014-01-04 13:05:16 -0600 received badge  Popular Question (source)
2014-01-02 02:41:04 -0600 asked a question 2 node Xenserver Openstack installation using RDO with Neutron OVS GRE

I am facing problems for some days now to get Xenserver Openstack installation working with Neutron OVS GRE.

I need clear instructions to create a 2 node Openstack Installation (one controller + compute + neutron, and the other compute node) on Xenserver 6.2 using RDO, along with Neutron OVS with GRE. So far, the instructions I found are few and have not worked. Any source of information would do. I am ready to try something other than RDO.

2013-12-17 02:29:21 -0600 commented answer How to safely clear Ceilometer database?

Using the above links, the following worked for me: Add line: vi /etc/mongod.conf smallfiles = true Add line: vi /etc/init.d/mongod MONGODB_OPTIONS="--nojournal"

2013-12-17 02:27:23 -0600 commented question What do I need to do to get a clean install of RDO Havana

Maybe you can start with giving us the error messages you get on the console when packstack fails. There will be additional information about log files in one of those messages.

2013-12-13 04:26:09 -0600 received badge  Enthusiast
2013-12-09 08:46:49 -0600 received badge  Teacher (source)
2013-12-09 08:46:49 -0600 received badge  Necromancer (source)
2013-12-09 06:23:58 -0600 commented answer RDO/Havana: Can't resize instance

Your previous answer was for someone else's question. I have +1 ed it, but could not accept it as an answer. I have +1 ed your comment here instead.

2013-12-09 06:15:55 -0600 commented answer RDO/Havana: Can't resize instance

You are absolutely right. I missed making the same change on the compute node. Thanks a million.

2013-12-09 04:28:57 -0600 answered a question RDO/Havana: Can't resize instance

I also have RDO/Havana on 64 bit machines with a controller node and a compute node. I tried resizing an instance, in both Running and Shutdown state. I tried this from the cli as well as horizon. Both times, I was able to locate the error in the logs:

"MigrationError(msg)\\n", "updated_at": "2013-12-09T10:13:34.773808", "id": 417, "finish_time": "2013-12-09T10:13:39.542276", "result": "Error", "request_id": "req-b0a0600d-7695-4661-b471-be23f2b8d064", "message": "destination same as source!","

As per some forums, I tried setting the following in /etc/nova/nova.conf and restarting nova services,

allow_resize_to_same_host=true
allow_migrate_to_same_host=true

but no luck. The instance goes into ERROR state, but loses it once I reboot it.

Does KVM support Resize? I can see from Hypervisor Support Matrix that KVM on x86 does support resize, but what about 64 bit machines? What am I missing in the configuration if Resize operation is supported?

Thanks.

2013-12-06 01:26:40 -0600 commented question why does cirros test image fail to launch?

Openstack works with Python 2.6 onwards. Is this issue resolved?

2013-12-06 01:18:10 -0600 answered a question Cannot access instance

Do you still have this problem?

When I had a similar error, I tried stopping iptables on both Controller and Compute node. It made the problem disappear. Later, I used RDO packstack to install Controller and Compute nodes, which creates appropriate iptables rules for access to various Openstack components.

service iptables stop

If this does not work, try disabling selinux in the file /etc/sysconfig/selinux.

2013-12-04 23:16:10 -0600 answered a question How to install havana on a single Node?

Easiest solution is using http://openstack.redhat.com/Quickstart (RDO Packstack). To control the components to be installed, generate an answers file, change the values, and run packstack with the answers file.

packstack --gen-answer-file my_answers.txt vi my_answers.txt packstack --answer-file=my_answers.txt

2013-12-04 07:00:11 -0600 received badge  Editor (source)
2013-12-04 06:48:23 -0600 answered a question How to avoid FAIL with user credentials when starting Autoscaling Template

I am not sure of the user privilege problem, as we create stacks from horizon. We log in using the tenant specific user, and face no issues with permissions.

If you can login to the VM created, please check the log at /var/log/heat-provision.log. This should give you an idea as to what happened. Usually, there is some error with one of the bash commands provided. Increasing the Timeout can also help. For example:

ElasticLoadBalancer: Type: AWS::ElasticLoadBalancing::LoadBalancer Properties: AvailabilityZones: {'Fn::GetAZs': ''} Listeners: - {LoadBalancerPort: '80', InstancePort: '80', Protocol: HTTP} HealthCheck: {Target: 'HTTP:80/', HealthyThreshold: '3', UnhealthyThreshold: '5', Interval: '30', Timeout: '5'}

2013-12-04 00:38:03 -0600 received badge  Supporter (source)