November 2021 - openstack-discuss

[horizon][dev]Handle multiple login sessions from same user in Horizon
by Luz de Avila, Arthur 01 Nov '21

01 Nov '21

Hi everyone, In order to improve the system my colleagues and I would like to bring up a new feature to Horizon. We found out that an user is able to login in Horizon with the same credentials in multiple devices or/and browsers. This may be not very secure as the user can login in many different devices or/and browsers with the same credential. Thinking on that, we would like bring up more control to the admin of the system in a way that the admin can enable or disable the multiple login sessions according to the needs of the system. For a better follow up of this propose, a blueprint has been opened with more details about the idea and concepts of this and we would like the onion of the community whether this feature make sense to implement or not. The blueprint opened on launchpad: https://blueprints.launchpad.net/horizon/+spec/handle-multiple-login-sessio… Kind regards, Arthur Avila

2 1

[charms] Yoga PTG Summary
by Alex Kavanagh 01 Nov '21

01 Nov '21

Hi All Thanks to all who participated in the PTG discussions. We had some good sessions where we tackled thorny issues in the charms deployments of OpenStack. A brief set of highlights is: - The store holding all the charms (https://jaas.ai/) just happens to be migrating to a new home (https://charmhub.io/) within the timescales of the Yoga cycle and some of the changes in behaviour provide some challenges and also opportunities to both simplify the charms and provide a better operator experience. The discussions highlighted the pain points and the general direction that the migration could take around: - Charm upgrades - OpenStack upgrades - Series upgrades of the underlying operating system. - The charms deployment of OpenStack supports versions back to mitaka (pre yoga) and queens (yoga cycle onwards). This has recently presented challenges around Python 3 support, particularly with py35 being EOL September 2020, and with py36 (bionic LTS) EOL 23rd December 2021. We discussed various strategies and approaches around maintenance, security and support of queens onwards in terms of support and upgrades. - The charms team also maintains a CI infrastructure based on Jenkins and (more recently) self-hosted Zuul to test installation and upgrade of OpenStack solutions. We discussed the further migrations of services from Jenkins to Zuul and issues associated with it. - Discussions around building new charms and refreshing old charms for the new components and features in the previous xena cycle and oncoming yoga cycle. Our raw notes are here: https://etherpad.opendev.org/p/charms-yoga-ptg Thank you again for all the participation in setting the direction and priorities for the yoga cycle. -- Alex Kavanagh - PTL Yoga

1 0

Openstack ansible VS kolla ansible
by A Monster 01 Nov '21

01 Nov '21

Openstack-ansible uses LXC containers to deploy openstack services , while Kolla uses docker containers instead, which of these two deployment tools should I use for an Openstack deployment, and what are the differences between them.

3 2

is OVS+DPDK useful for general purpose workload
by Satish Patel 01 Nov '21

01 Nov '21

Folks, I have deployed openstack and configured OVS-DPDK on compute nodes for high performance networking. My workload is general purpose workload like running haproxy, mysql, apache and XMPP etc. When I did load testing I found performance was average and after 200kpps packet rate I noticed packet drops. I heard and read that DPDK can handle millions of packets but in my case its not true. I am using virtio-net in guest vm which processes packets in the kernel so I believe my bottleneck is my guest VM. I don't have any guest based DPDK applications like testpmd etc. does that mean OVS+DPDK isn't useful for my cloud? How do I take advantage of OVS+DPDK with general purpose workload? Maybe I have the wrong understanding about DPDK so please help me :) Thanks ~S

2 2

Re: Using ceph for openstack storage
by A Monster 01 Nov '21

01 Nov '21

Thank you for your response. Sadly, I'm talking about actual production, but I'm very limited in terms of hardware. I was thinking about using RAID for controller node as data redundancy, because I had the idea of maximizing the number of nova compute nodes, So basically i thought off using a controller with the following services ( Nova, Neutron, Keystone, Horizon, Glance, Cinder and Swift). Following the configuration you suggested, I would have : - 3 Controllers that are also Ceph Monitors - 9 Nova compute nodes and Ceph OSDs My questions are : - having multiple Ceph monitors is for the sake of redundancy or does it have a performance goal ? - Combining Ceph OSD and Nova compute wouldn't have performance drawbacks or damage the integrity of data stored in each node. wouldn't it be better is this case to use two separate servers for swift and glance and use RAID for data redundancy instead of using Ceph SDS. Thank you very much for you time.

3 2

Re: Router interfaces are down
by Slawek Kaplonski 01 Nov '21

01 Nov '21

Hi, Please don't drop ML from the thread. You have to go to the node where Your router is hosted and investigate there in the agent's logs. If ports are DOWN, I would start with checking in the L2 agent logs (neutron-ovs-agent or linuxbridge-agent, idk what You are using exactly). If there is no any errors there, You can also check neutron-server logs, why ports aren't set to be UP as well as checking neutron-l3-agent logs on the node where router is hosted. On czwartek, 28 października 2021 18:41:29 CET Jibsan Joel Rosa Toirac wrote: > Here I let you some screenshots of my Network Topology: > > > > El jue, 28 oct 2021 a las 7:43, Jibsan Joel Rosa Toirac (<jibsan94(a)gmail.com>) > escribió: > > Yes it’s neutron router. Well the router is centralized. It is located > > between the subnets and the node, all the subnets will pass through the > > router to internet, but I don’t know what else check to set it up for. I’m > > using a friend’s Openstack instance on a server to check out if I’m missing > > something and both nodes are the same. I will send you a screenshot of my > > network Topology. > > > > Greetings > > > > On Thu, Oct 28, 2021 at 2:33 AM Slawek Kaplonski <skaplons(a)redhat.com> > > > > wrote: > >> Hi, > >> > >> On środa, 27 października 2021 21:39:55 CEST Jibsan Joel Rosa Toirac > >> > >> wrote: > >> > Hello, I'm trying to route all the requests from a private vlan to > >> > internet. I have a private network and all the Virtual Machines inside > >> > >> the > >> > >> > subnet config can do everything between they, but if I ping to > >> > >> Internet, it > >> > >> > doesn't work. > >> > > >> > When I see the router_external it says all the interfaces are DOWN. > >> > >> By router_external, You mean neutron router, right? > >> If so, what kind of router it is, centralized HA or non-HA, or maybe DVR? > >> Is router scheduled properly to some node? You can check that with > >> command > >> like "neutron l3-agent-list-hosting-router <router_id>". > >> > >> > I have search in everywhere but I can't find a solution for this. > >> > > >> > Thank you for your time > >> > >> -- > >> Slawek Kaplonski > >> Principal Software Engineer > >> Red Hat -- Slawek Kaplonski Principal Software Engineer Red Hat

1 0

[Diversity] Diversity and Inclusion Meeting Reminder
by Amy Marrich 31 Oct '21

31 Oct '21

The Diversity & Inclusion WG invites members of all OIF projects to attend our next meeting Monday November 1st, at 17:00 UTC in the #openinfra- diversity channel on OFTC. The agenda can be found at https://etherpad.openstack.org/p/diversity-wg-agenda. Please feel free to add any topics you wish to discuss at the meeting. Thanks, Amy (apotz)

1 0