June 2020 - openstack-discuss - lists.openstack.org

Delegating and routing an IPv6 prefix to an instance
by Christian Rohmann 29 Jun '20

29 Jun '20

Hello OpenStack-Discuss, I have a use case in which an instance / VM is hosting i.e. an OpenVPN gateway which shall be doing some routing / networking by itself. For that purpose one would like to have a global unique IPv6 prefix delegated and routed to it to, which it can in turn give out to its VPN clients. This can and should not be cut out of the on-link network that is provided by Neutron and used to connect the instance itself. If you look at https://community.openvpn.net/openvpn/wiki/IPv6, which has a section *Details: IPv6 routed block* explaining just how that is one intended approach on how to do it. I am now wondering if the existing DHCPv6 prefix delegation implemented in OpenStack is capable of providing a prefix to an instance. Digging a little into what can be found online I ran into this Etherpad doc https://etherpad.opendev.org/p/neutron-kilo-prefix-delegation (linked to on https://wiki.openstack.org/wiki/Neutron/IPv6/PrefixDelegation) There is a list of use-cases, the second one being exactly what I described above: > [...] > > Use cases: > > We need to allocate addresses to ports from an external or > providernetwork, and route them via Neutron routers. > > We wish to allocate whole prefixes to devices (and their specific > neutron port) on demand. A port must be authorised via the API for a > prefix. The prefix could be issued to the device via PD (since the > device has to discover the prefix it's been given). > > [...] But to my understanding the spec used to implement the current IPv6 networking and also prefix delegation mechanism, also mentioned this use case as an "limitation and future enhancement" - see: https://specs.openstack.org/openstack/neutron-specs/specs/liberty/ipv6-pref… Does anyone have any thoughts on this matter of dedicating a prefix and and routingits traffic to a VM, but not just a subnet? Regards Christian

4 8

[all][qa] devstack dropping python 2 support
by Ghanshyam Mann 29 Jun '20

29 Jun '20

Hello Everyone, Devstack has been default to python3 since ussuri release and Victoria cycle is time to drop the support of python2 completely. - https://review.opendev.org/#/c/737455/ Being ussuri as python3-only release and all grenade jobs has been migrated to python3 in ussuri, we are all good here but in case your project is still using the py2 jobs or replying on python2 support in devstack in any extend, do ping us on #openstack-qa channel. We will be keeping USE_PYTHON3 variable support till all consumer remove this, like storlets to adjust the requirement. NOTE: branchless repo keeping common jobs for master and stable branch gate with USE_PYTHON3=False need to keep setting this var for configuring the devstack with right python. For example, tempest-full job definition in tempest master. Such jobs should be run till stable/train only as ussuri onward we tests things on python3 only. -gmann

1 0

Monasca duplicate metrjcs
by junaid soomro 29 Jun '20

29 Jun '20

Hello everyone, I am new to this. my scenario is when i try to get io.read_bytes metric from monasca api it gives me more than one metrics of a singke hostname. Although device is different vda vdb etc. This is fine but when i remove the volume from the vm. Monasca api still sbows its metrics. Ideally it should remove the vdb because i removed it but it does not. why does this happen? Thank you

2 1

RE: [neutron] DVR / IPv6 on provider network instead?
by Eric K. Miller 28 Jun '20

28 Jun '20

> If so, is there a possibility of using a provider network that is connected to all compute nodes where IPv6 subnets are issued to tenants from a subnet pool, with traffic being routed directly to an external router (not a Neutron router) using Linux Bridge instead of OVS? Yet, still use port security? > Just trying to figure out the best way to support IPv6 without forwarding all traffic through a single network node, while using DVR for IPv4. I believe I have a solution, which works for IPv4, where public IP addresses are assigned to a VM's port without the use of floating IPs. I think this approach can work for IPv6, with an upstream DHCPv6 server. My IPv4 approach was to create an internal VLAN on our firewalls, where a public subnet is assigned (statically routed), and a gateway IP is assigned to the firewall. This VLAN is then assigned to a provider network. An OpenStack subnet is created on the provider network for the public subnet range with an allocation range associated with the IPs we want to hand out to VMs, with DHCP enabled. The subnet's gateway property is set to "None", along with DNS servers set, and a "host-route" is added for a default gateway, using our firewall's IP as the gateway. When a server is created on this network, an IP is allocated from the public subnet, the gateway is assigned to our firewall's IP, DNS is set correctly, and security groups work fine on the port. Best of all, traffic is sent/received from the host, not through the network node, so this is a reasonably scalable solution. Eric

1 0

[cyborg][nova] Support flexible use scenario for Mdev(such as vGPU)
by Feng, Shaohe 28 Jun '20

28 Jun '20

Hi all, Currently openstack support vGPU as follow: https://docs.openstack.org/nova/latest/admin/virtual-gpu.html In order to support it, admin should plan ahead and configure the vGPU before deployable as follow: https://docs.openstack.org/nova/latest/configuration/config.html#devices.en… This is very inconvenient for the administrator, this method has a limitation that a same PCI address does not provide two different types. Cyborg as an accelerator management tool is more suitable for mdev device management. One solution as follow： Firstly, we need a vender driver(this can be a plugin), it is used to discovery its special devices and report them to placement for schedule. The difference from the current implementation is that: 1. report the mdev_supported_types as traits to resource provider. How to discover a GPU type: $ ls /sys/class/mdev_bus/*/mdev_supported_types /sys/class/mdev_bus/0000:84:00.0/mdev_supported_types: nvidia-35 nvidia-36 nvidia-37 nvidia-38 nvidia-39 nvidia-40 nvidia-41 nvidia-42 nvidia-43 nvidia-44 nvidia-45 so here we report nvidia-3*, nvidia-4* as traits to resource provider. 2. Report the number of allocable resources instead of vGPU unit numbers to resource provider inventory Example for the NVidia V100 PCIe card (one GPU per board) : Virtual GPU Type Frame Buffer (Gbytes) Maximum vGPUs per GPU Maximum vGPUs per Board V100D-32Q 32 1 1 V100D-16Q 16 2 2 V100D-8Q 8 4 4 V100D-4Q 4 8 8 V100D-2Q 2 16 16 V100D-1Q 1 32 32 so here we report 32G Buffers(an example, maybe other resources) to resource provider inventory 3. driver should also support a function to create certain mdev type, such as (V100D-1Q, V100D-2Q,) Secondly, we need a mdev extend ARQ(it can be a plugin): Here is an example for fpga ext arq: https://review.opendev.org/#/c/681005/26/cyborg/objects/extarq/fpga_ext_arq… The difference is that, we replace the _do_programming to _do_create_mdev For _do_programming, it is used to create a new FPGA function. For _do_create_mdev, it is used to create a new type mdev, it will call the implementation function in vendor driver. At last we need to support a mdev handler for xml generation in nova, we can refer to the cyborg PCI handler in nova So after the above changes： Admin can create different SLA devices profiles such as: {“name”: “Gold_vGPU”, “groups”: [ {“resources:vGPU_BUFFERS”: “16”, “traits: V100D-16Q,”: “required”, }] } And {“name”: “Iron_vGPU”, “groups”: [ {“resources:vGPU_BUFFERS”: “1”, “traits: V100D-1Q,”: “required”, }] } Then a tenant can use Gold_vGPU to create with a VM with V100D-16Q vGPU And another tenant can use Iron_vGPU to create with a VM with V100D-1Q vGPU When ARQ binding during the VM creating, the Cyborg will call the vendor driver to create expected mdev vGPU. And these 2 mdev vGPU can be on same physical GPU card. The mdev extend ARQ and vendor driver can be plugin, they are loose couple with the upstream code. So the downstream can get the upstream code to customize the own mdev extend ARQ and vendor driver. Here vGPU is just an example, it can be other mdev devices. BR Shaohe Feng

3 11

RE: [neutron] DVR / IPv6 on provider network instead?
by Eric K. Miller 28 Jun '20

28 Jun '20

> Also, unrelated, but hopefully a quick question... is the "internal" or "external" label on a network just used for filtering lists, such as for "openstack network list --external"? or does it change the behavior of anything? Regarding my second question, I found some information that indicates --external will share the network with all projects. Whereas the --shared option is only used for sharing a network with a specified list of projects using RBAC rules. It also looks like networks marked as external, not internal, can be specified for the --external-gateway network when creating/updating a router. This has some good info too: https://superuser.openstack.org/articles/tenant-networks-vs-provider-networ… Eric

1 0

[ops][cinder] removal of brocade fczm driver
by Brian Rosmaita 26 Jun '20

26 Jun '20

Brocade no longer supports their cinder Fibre Channel Zone Manager driver [0]. It was marked 'unsupported' in Ussuri and deprecated for removal during the Victoria development cycle. (We just discovered this week that it is in fact unusable in Ussuri as it does not run in Python 3. It does pass all unit tests in Python 3, though.) Its removal will leave cinder with only one FCZM driver, so there has been some interest among Cinder project members in fixing the driver so it will at least run in Python 3.6 and 3.8. However, we don't know at the moment how much effort that will take. Further, without a third-party CI continually testing this driver, we cannot guarantee that once fixed, it will continue to function correctly. (No human can guarantee that any software of sufficient complexity functions correctly, but you know what I mean.) The point of this email is threefold: 1. If nobody is using the Brocade FCZM driver, then maybe we should just go ahead and remove it. So if you are using it, please let us know. 2. According to [0], for OpenStack users after Train, "Brocade encourages customers to perform SAN Zoning operations using other available tools or APIs." So if you are a Brocade user who has successfully configured another solution to replace the Brocade FCZM, the Cinder team (and the wider OpenStack community) would be interested in hearing about it. 3. If you are a consumer of the current cinder Brocade FCZM and are interested in its continued existence, the Cinder team would be interested in discussing with you how we could set up a third-party CI for it. As I said earlier, without the third-party CI, we really have no idea whether this thing works or not. I've already posted a patch to remove the Brocade FCZM driver [1]. (Just so you know we're serious about this.) At our R-16 mid-cycle meeting earlier this week, the Cinder team agreed to make a final decision at our next mid-cycle at week R-9. (Yes, we have two mid-cycles and neither is at the middle of the cycle -- we're already meeting in virtual space, so we also meet in virtual time.) So please communicate your thoughts to us before week R-9 by replying to this email on the mailing list. cheers, brian [0] https://docs.broadcom.com/doc/12397527 [1] https://review.opendev.org/#/c/738148/

1 0

[all][interop] Tasks completed and help for Interop Survey questions
by prakash RAMCHANDRAN 26 Jun '20

26 Jun '20

Hi all, It was bit odd to use meetpad instead of irc today, still need to understand how to share multiple screens or desktop on the meetpad.Any help is welcome as hat the procedure will be. Plus apology to Mark for late update on irc to meetpad. Weekly we will be using - (https://etherpad.opendev.org/p/interop) Interop Working Group - Weekly Friday 10-11 AM or UTC 17-18 Lnink: https://meetpad.opendev.org/Interop-WG-weekly-meeting But communication was really better than irc with chat plus verbal+video talk. How to add ircbot to meetpad any suggestions? Following is action Items to follow up. AI-01- GM to update/close? launchpad item - https://blueprints.launchpad.net/refstack/+spec/interop-2020.06With following two references to storyboard for all API guidelines for 2020.06.json.https://storyboard.openstack.org/#!/story/2007509https://story… AI-02 - Survey Questions please add what you seek for your projects? - https://storyboard.openstack.org/#!/story/2007511 6/26/2020 - AT todays Interop call it was decided to add following questions and is being circulated for updates via openstack-discuss list.level 1 Q &A1.Do you use OpenStack Marketplace data for your decision on use of OpenStack Cloud?2.Do you look at Interoperability reports available to you on different offerings in Marketplace portal?3. If you offer OpenStack Cloud do you use the OpenStack logo programs.?4. If you are using Object Storage for your cloud do you use Swift? What would you like to see improve in current Marketplace Logo Program?[1] Do you want to add Interop for kubernetes API conformance for OpenStack and Open Infrastructure Projects?[2] Do you want to add Interop for bare metal service?[3][4] (Add above your suggestions)level 2 Q&A1. Do you use data on add-on programs in the OpenStack Marketplace?1.1 If yes which ones? AI-03 - (for Core + Add-ons for OpenStack Integrated Projects) - Can you volunteer to add tag assert:supports-api-interoperability to your projects?Any comments on what Projects need to do to claim this based on 2020.06 - interop - https://opendev.org/osf/interop/src/branch/master/2020.06.json Can someone take a crack at documenting how to for one project like nova compute api does now? AI-04 - Seeking volunteers to maintain - Refstack - https://opendev.org/osf/refstack AI-05 - Can someone volunteer to present at next Interop meeting Friday 3rd at 17 UTC (10 AM PDT) - https://github.com/cncf/k8s-conformance Any Other suggestions for future events? ThanksInterop team: Prakash/Mark/Ghanshyam/Arkady

1 0

[openstack-discuss][airship-discuss][openstack-helm] Ceph Bluestore
by Roman Gorshunov 26 Jun '20

26 Jun '20

Hello Dominik, Thanks for the question. Please, let me redirect it to the OpenStack-Helm project mailing list, as Airship only consumes Ceph charts from openstack-helm-infra repository (previously from openstack-helm repo). Technically I see that ceph-osd in OpenStack-Helm supports Bluestore [0], so if you would want to deploy with Bluestore, you would need to change Armada documents (e.g. [1][2]) to use Bluestore. [0] https://opendev.org/openstack/openstack-helm-infra/src/branch/master/ceph-o… <https://opendev.org/openstack/openstack-helm-infra/src/branch/master/ceph-o…> [1] https://opendev.org/airship/treasuremap/src/branch/master/site/seaworthy/so… <https://opendev.org/airship/treasuremap/src/branch/master/site/seaworthy/so…> [2] https://opendev.org/airship/treasuremap/src/branch/master/site/seaworthy/so… <https://opendev.org/airship/treasuremap/src/branch/master/site/seaworthy/so…> Best regards, Roman Gorshunov > On 25 Jun 2020, at 17:36, Dominik Bender <db(a)db-itsolutions.com> wrote: > > Hello everyone, > > I need some help with Ceph configuration. What changes need to be made in the charts to use bluestore? > In general I would still be interested in why bluestore is not yet standard in treasuremap? > > Thanks in advance > > -- > Dominik Bender > > > _______________________________________________ > Airship-discuss mailing list > Airship-discuss(a)lists.airshipit.org > http://lists.airshipit.org/cgi-bin/mailman/listinfo/airship-discuss

1 0

OpenStack on CentOS 8 VM using packstack
by Pradeep Antil 26 Jun '20

26 Jun '20

Hi Techies, I am trying to deploy the latest Openstack "ussuri" on CentOS 8 VM using packstack. I have successfully installed it but the issue is that i want to i want to add my interface "enp0s3" to ovs bridge br-ex. On CentOS 7 we can do this via ifcfg-* files using network-scripts. But i am not sure how this can be done on CentOS 8 VM as in this release network is controlled via network manager. On br-ex i will attach an external provider network or a flat network. [image: image.png] I have referred the below steps for the deployment. dnf update -y dnf config-manager --enable PowerTools dnf install -y centos-release-openstack-ussuri dnf update -y dnf install -y openstack-packstack reboot packstack --gen-answer-file=/root/answer.txt Edit answer file CONFIG_NEUTRON_ML2_TYPE_DRIVERS=vxlan,flat CONFIG_NEUTRON_ML2_TENANT_NETWORK_TYPES=vxlan CONFIG_PROVISION_DEMO=n CONFIG_KEYSTONE_ADMIN_PW=P@ssw0rd packstack --answer-file=/root/answer.txt Following is the output of ovs command and interfaces on my centos 8 system [image: image.png] -- Best Regards Pradeep Kumar

2 1