First time here? Check out the FAQ!
2020-09-02 01:30:28 -0600 | commented question | How to setup initial networking (Lan/Internet) after fresh installation The "external" network in a default Packstack installation is an isolated test network that is not connected to anything. For a truly external network, try https://www.rdoproject.org/networking.... |
2020-09-01 01:20:59 -0600 | commented question | Senlin Installation using Juju and Senlin from Ubuntu Repo If you opt for installation from the Git source, install it on a controller. However, in principle it should probably work no matter where it is installed, even on a node outside your current OpenStack cluster. |
2020-09-01 01:13:10 -0600 | commented question | DHCP not working FLAT network What do you mean by "DHCP doesn't work"? Did you configure DHCP, and how? Which router do you try to ping? Are you trying to access static or floating IPs? Can you access those IP addresses from the controller? Have you configured instances' security groups? |
2020-08-30 09:33:48 -0600 | commented question | openstack packstack ussri installation fails Centos 8 Is this all from the same packstack invocation? Also, the first error mentions two log files under /var/tmp/packstack/20200824-071619-qxdk96sr. Did you check them? |
2020-08-30 09:32:08 -0600 | commented question | openstack packstack ussri installation fails Centos 8 This is very confusing, because you have several errors. A timeout when running nova manage db-sync. A Connection Refused when trying to get access to Keystone. And now Keystone is running but since project services doesn't exist (perhaps a consequence of not getting access to Keystone earlier). |
2020-08-29 19:39:53 -0600 | answered a question | Multi-Node Lab tutorial Yes, you need to create the worker nodes and configure them as suggested at https://docs.openstack.org/devstack/l.... Gateway is the router in your network. If you don't know the meaning of the gateway clause, I suggest you study networking in general and Linux networking in particular before you set up an OpenStack cluster. Without networking skills, you will have plenty of problems. By the way, the network setup in the instructions only works for Ubuntu 16 and earlier. Ubuntu 18 and later uses a different configuration tool named netplan. |
2020-08-29 05:39:37 -0600 | commented question | openstack packstack ussri installation fails Centos 8 It could be that Keystone takes too long to start, i.e. the problem might be caused by a race condition. If so, I am not sure how to fix that. |
2020-08-29 05:37:50 -0600 | commented question | openstack packstack ussri installation fails Centos 8 By the way, I don't see ProjectNotFound in the output that you posted. |
2020-08-29 05:33:33 -0600 | commented question | openstack packstack ussri installation fails Centos 8 Well, is it running? Of course it's supposed to be running, since the command that fails is |
2020-08-28 14:02:12 -0600 | commented question | openstack packstack ussri installation fails Centos 8 The error indicates that Keystone is not running. Does the server have enough RAM? Does the Keystone log or the kernel log (journalctl -k) contain information about Keystone crashing? |
2020-08-28 04:17:58 -0600 | commented question | Integrating Mistral with remote Ussuri controller I don't pretend I know anything about VNFs, Tacker and Mistral, but very generally it might help if you shared the command you issue to populate the DB and the corresponding error message. |
2020-08-25 17:38:46 -0600 | commented answer | Chrony on controller node doesn't list the compute node This line tells the controller to use itself as NTP source. I don't know if this is possible at all, or if other configuration settings are required for this scenario. Perhaps you need the local directive to synchronize with the local clock. |
2020-08-24 22:06:42 -0600 | answered a question | Chrony on controller node doesn't list the compute node You didn't configure the compute node as an NTP server (there is no allow clause). Therefore, it doesn't appear in the controller's time sources. In fact, you didn't configure any time source on the controller (no server or pool clause). Furthermore, this line in the controller configuration doesn't look correct (I am not sure): As a result, I guess the compute node doesn't get NTP services from the controller. Try, for example, 192.168.0.0/24, or allow all. I suppose that 192.168.0.11 is the controller, correct? |
2020-08-21 08:15:41 -0600 | answered a question | how swift to be the backend of glance to store image
|
2020-08-20 07:07:04 -0600 | commented question | stuck openstack instance From what images did you launch these instances? Are they configured in any way? Have you tried other images? Can you ping them? Can you generate a console log? |
2020-08-20 03:58:43 -0600 | commented question | Give VM access to outside world (question) And don't mess with netfilter rules (iptables). |
2020-08-20 03:58:07 -0600 | commented question | Give VM access to outside world (question) I am not too knowlegeable about it, but here are several options how to configure DNS: https://docs.openstack.org/neutron/ro.... |
2020-08-20 03:56:20 -0600 | commented question | Give VM access to outside world (question) ping 1.1.1.1 succeeds. Congratulations, you have internet access! What you don't have is name resolution. I think there are parameters for the DHCP service that help with that, but for a quick fix, you could add 1.1.1.1 or your local nameserver's address to /etc/resolv.conf. |
2020-08-19 17:51:30 -0600 | commented question | Give VM access to outside world (question) Can you show me:
Can you ping 1.1.1.1 from an instance, and from the host? |
2020-08-18 18:55:11 -0600 | commented question | Problems with openstack newton on CentOS 7 I'd also like to see the details of the external network. |
2020-08-18 18:53:41 -0600 | commented question | Problems with openstack newton on CentOS 7 The information you provided contains nothing that raises my suspicion, but you say that instances can't connect to their router, so that I think we should look into the router. Are you able to show me the IP addresses and routing table in the router's network namespace? |
2020-08-18 18:51:54 -0600 | commented question | Problems with openstack newton on CentOS 7 How did you set up the cloud? When you try to connect from the host to the instances, I suppose you use their floating IPs, correct? Can you show details of the Neutron router that connects private-01 and public-03? |
2020-08-18 18:43:48 -0600 | edited question | Problems with openstack newton on CentOS 7 I recently did all in one node Openstack Newton setup on CentOS 7 box. It went with much less hussle. But I am facing the networking issue of connecting to the instances which are on virtual network with the controller node. 1. I cannot ping the nodes (172.17.y.y) in private virtual network from the host node(192.168.x.x). I already added entries into static routing table but its not making any difference. Tried adding additonal roytes as part of the private virtual network subnet from Openstack dashboard, still the problem persists.
It would be really great if someone can help in resolving any of the above issues. Thanks in Advance ! # rpm -qa | grep newtoncentos-release-openstack-newton-1-2.el7.noarch Virtual Network setup : #### ip a (more) |
2020-08-18 18:28:20 -0600 | commented question | Give VM access to outside world (question) There is no router at all. How can these two instances connect to the lab environment? Create a router between shared and public and try again. |
2020-08-18 08:53:56 -0600 | commented question | SDN for Openstack See https://www.openstack.org/analytics and select Deployment Decisions. One of the graphs shows the number of deployments per Neutron plugin. This should help answering the second question. The term "SDN" is a bit vague. I have seen it applied to Neutron itself. You may want to be more specific. |
2020-08-18 08:45:33 -0600 | answered a question | upload/manage files in the object store There is no difference whether you do it from an OpenStack instance or from another machine. See https://docs.openstack.org/python-ope... and https://docs.openstack.org/python-swi.... Of course, you can also use the object store API or write programs using the OpenStack SDK. |
2020-08-18 00:25:49 -0600 | commented question | Give VM access to outside world (question) I don't know what else to try, since I don't know what you have tried. Therefore my question: What do you try to connect to the internet? |
2020-08-18 00:23:12 -0600 | commented question | Problems with openstack newton on CentOS 7 The Neutron install guide shows how to set up an external network. |
2020-08-18 00:21:40 -0600 | commented question | Problems with openstack newton on CentOS 7 Why Newton? It's ancient. How did you set up the cloud, and can you provide details of the virtual network setup? |
2020-08-16 19:49:02 -0600 | commented question | Give VM access to outside world (question) You are right: https://imgur.com/7aHULtV shows that the interface is connected to the bridge, but in order to communicate with the world outside the cloud, the bridge must have an IP address that allows it to do that. Which it didn't in the previous setup. |
2020-08-16 18:46:22 -0600 | commented question | Give VM access to outside world (question) If your cloud is now connected to the local network but can't reach the internet, my guess is that something needs to be configured in the local network. What do you try to connect to the internet? |
2020-08-16 00:49:56 -0600 | commented question | Give VM access to outside world (question) To confirm that enp2s0 is plugged into br-ex, also run ovs-vsctl show. |
2020-08-16 00:47:57 -0600 | commented question | Give VM access to outside world (question) Not sure if you can repair the current config. Perhaps by adding an external address to br-ex, then creating a second external network (see https://docs.openstack.org/neutron/la...). |
2020-08-16 00:43:13 -0600 | commented question | Give VM access to outside world (question) br-ex has no outside network address, which explains why instances aren't connected to the external network. I guess that the instructions are not complete. Try setting FLOATING_RANGE, PUBLIC_NETWORK_GATEWAY and Q_FLOATING_ALLOCATION_POOL as well. I believe GATEWAY is the desired address of br-ex. |
2020-08-15 19:24:34 -0600 | commented question | Give VM access to outside world (question) I guess you did https://docs.openstack.org/devstack/l..., correct? Please add local.conf to your question, and the output of ip a show br-ex. How are your instances networked? What do you mean by "only made it visible for the host"? |