January 2019 - Starlingx-discuss

[Starlingx-discuss] [build] go packages, version 2
by McKenna, Jason 17 Jan '19

17 Jan '19

Hi build team, At the previous build meeting I had identified an issue with the way some go based packages were being built be built (they required internet access), and promised I'd update the mailing list on a potential way forward that we were prototyping. Some preliminary points: - go usually attempts to resolve dependencies at build time, by going out to the internet and fetching stuff (like dependency source code) using the "go get" command - Sometimes the stuff fetched by "go get" isn't appropriate (i.e. "go get" fetches the latest version, but deprecated APIs may have been removed, etc) - Different versions of go packages may require different versions of dependencies - We want builds to be reproducible without unexpected code changes (i.e. we want to know what we're compiling in) - Some people build in environments where they don't have Internet access The initial solution (which didn't take into account the Internet access problem) was to use "dep". "dep" is an external tool which was an "official experiment" of the go project. Rather than fetch the latest dependencies from the internet (like "go get"), it allowed specific revisions of dependencies to be captured. "dep" fetched those versions from the Internet. This solved the deprecated API issue, the reproducible build issue, and the issue of not using rpms for the dependencies. However, if someone was attempting to build in an internet-less context, the system would fail. Enter this second revision. The dependency packages are now downloaded at download-mirrors.sh time as tarballs. The tarballs are produced as of a specific commit for each dependency. This allows us to hit all our bullet points - code is snapshotted, reproducible, we don't end up having to create a bunch of new rpms with dependency source code and potential version conflicts, and it requires no internet access (other than at download-mirrors.sh time) Jerry has posted a preview code review showing his work using this mechanism. I've marked the reviews as workflow -1 to give the build team a chance to see the mechanism. https://review.openstack.org/#/c/631001/ https://review.openstack.org/#/c/631002/ -Jason

3 4

[Starlingx-discuss] Installation on a specific raid configuration?
by Marcel Schaible 17 Jan '19

17 Jan '19

Hi, I am trying to install the 20190109 iso on a raid configuration with two nvme disks. The installer by default creates the following disk layout and does not obye an existing partioning scheme: # Start End Size Type Name 1 2048 616447 300M EFI System EFI System Partition 2 616448 1640447 500M Microsoft basic 3 1640448 42600447 19.5G Microsoft basic 4 42600448 430573567 185G Linux LVM Questions: For waht are these partition #2 and #3 are used? Actually we would like to have for a fail over scenario the following layout: # Start End Size Type Name 1 2048 616447 300M EFI System EFI System Partition 2 616448 ...... 950G Raid mirrored to second nvme1n1p2 3 ....... ...... 32G swap Is there a way to tell the installer this configuartion? Thanks Marcel

3 4

[Starlingx-discuss] How to get access to Starlingx
by bharath thiruveedula 17 Jan '19

17 Jan '19

Hi, I am trying to install Starlingx, but due to resource constraints. couldn't achieve it. Can we install Starlingx on a 16GB machine? If not, is there any way to explore the features of starlingx like accessing public installation of Starlingx? Best Regards Bharath T

1 0

[Starlingx-discuss] cpu mode
by von Hoesslin, Volker 17 Jan '19

17 Jan '19

hi, my setup has two computes nodes, every node has a dual AMD EPYC 7601 CPU config. how can i bring all the CPU features (AES, SSSE3, ...) to the guest VMs. i have tryed with some flavor-metadata but nothing realy helps, the VMs getting just a little subset of cpu-features. some investigations to the kvm-settings hit me to the facts that my nova config has "cpu_model=none" !? how can i fix that and bring my AMD EPIC CPU to my nova-config?! here is the host /proc/cpuinfo processor : 127 vendor_id : AuthenticAMD cpu family : 23 model : 1 model name : AMD EPYC 7601 32-Core Processor stepping : 2 microcode : 0x8001227 cpu MHz : 1200.000 cache size : 512 KB physical id : 1 siblings : 64 core id : 31 cpu cores : 32 apicid : 127 initial apicid : 127 fpu : yes fpu_exception : yes cpuid level : 13 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc art rep_good nopl nonstop_tsc extd_apicid amd_dcm aperfmperf eagerfpu pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_l2 cpb hw_pstate retpoline_amd ssbd ibpb vmmcall fsgsbase bmi1 avx2 smep bmi2 rdseed adx smap clflushopt sha_ni xsaveopt xsavec xgetbv1 clzero irperf xsaveerptr arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif overflow_recov succor smca bogomips : 4400.08 TLB size : 2560 4K pages clflush size : 64 cache_alignment : 64 address sizes : 48 bits physical, 48 bits virtual power management: ts ttp tm hwpstate cpb eff_freq_ro [13] [14] greez & thx, volker...

1 0

[Starlingx-discuss] Sanity Test - ISO 20190116
by Alonso, Juan Carlos 16 Jan '19

16 Jan '19

Status of the Sanity Test for last CENGN ISO: bootimage.iso from 2019-Jan-16 (link<http://mirror.starlingx.cengn.ca/mirror/starlingx/master/centos/20190116T06…>) Sanity Test is executed in a Virtual Environment Status: GREEN Simplex Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 25 TCs [PASS] TOTAL: [ 30 TCs PASS ] Duplex Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 26 TCs [PASS] TOTAL: [ 31 TCs PASS ] Multinode Controller Storage Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 26 TCs [PASS] TOTAL: [ 31 TCs PASS ] Multinode Dedicated Storage Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 26 TCs [PASS] TOTAL: [ 31 TCs PASS ] ------------------------------------------------------------------ Regards. Juan Carlos Alonso

1 0

Re: [Starlingx-discuss] Approach for SB 2004710 support to access docker images via proxy?
by Peters, Matt 16 Jan '19

16 Jan '19

Hello, I also wanted to bring to your attention the following spec and storyboard. These outline changes forthcoming for replacing config_controller. Therefore, any changes that are introduce for config_controller should consider the impacts to this development. https://review.openstack.org/#/c/629581/ https://storyboard.openstack.org/#!/story/2004695 From: "Qi, Mingyuan" <mingyuan.qi(a)intel.com> Date: Monday, January 14, 2019 at 9:29 PM To: Barton Wensley <Barton.Wensley(a)windriver.com>, "Miller, Frank" <Frank.Miller(a)windriver.com>, "Kung, John" <John.Kung(a)windriver.com> Cc: Chris Friesen <Chris.Friesen(a)windriver.com>, "Church, Robert" <Robert.Church(a)windriver.com>, Brent Rowsell <Brent.Rowsell(a)windriver.com>, "Xie, Cindy" <cindy.xie(a)intel.com>, "Penney, Don" <Don.Penney(a)windriver.com>, "Peters, Matt" <Matt.Peters(a)windriver.com> Subject: RE: Approach for SB 2004710 support to access docker images via proxy? Thanks Bart, I agree with the proxy part. As for alternate docker registry I have different thinking. The additional docker registry(or called registry mirror) is to accelerate docker image pulling from internet, not only for openstack-helm, but also for k8s/armada image. It’s slightly different from local docker registry (say controller_address:9001) while local registry is more like a cache for controller after pulling. If the registry mirror is needed for k8s/armada, it has to be set in config_controller as well. Meanwhile the registry mirror address may be in no_proxy if it’s host is within the same LAN as controller. For the proxy SB, once which table is confirmed to store proxy info, I will finish the code for review. Thanks, Mingyuan From: Wensley, Barton [mailto:Barton.Wensley@windriver.com] Sent: Tuesday, January 15, 2019 4:58 To: Miller, Frank <Frank.Miller(a)windriver.com>; Qi, Mingyuan <mingyuan.qi(a)intel.com>; Kung, John <John.Kung(a)windriver.com> Cc: Friesen, Chris <Chris.Friesen(a)windriver.com>; Church, Robert <Robert.Church(a)windriver.com>; Rowsell, Brent <Brent.Rowsell(a)windriver.com>; Xie, Cindy <cindy.xie(a)intel.com>; Penney, Don <Don.Penney(a)windriver.com>; Peters, Matt <Matt.Peters(a)windriver.com> Subject: RE: Approach for SB 2004710 support to access docker images via proxy? Mingyuan, As far as the input of the new configuration goes, we typically only add configuration questions to config_controller if the data is required for the execution of config_controller. In your case, I expect the proxy configuration would be required by config_controller (in order to download the kubernetes images), but the alternate docker registry wouldn’t be required until the stx-openstack application was required. So for the proxy configuration, we would probably add a new set of questions something like this: Kubernetes Configuration: ------------------------- Configure http proxy [y/N]: y HTTP proxy URL: http://proxy.example.com:80 HTTPS proxy URL: https://proxy.example.com:443 I suspect that the NO_PROXY for the service config file can be calculated ourselves - I assume it will just have the IP of our internal docker registry? The http proxy config would be system wide configuration value, so it wouldn’t belong in the host table. John Kung or Matt Peters can comment on which sysinv table you should use. The configuration of the http-proxy.conf file can be done in the docker.pp manifest as you suggested below. This needs to be done on both controllers. You need input from Bob Church for the configuration of the alternate docker registry. I think the registry is currently hardcoded in sysinv. I suspect we will want a new sysinv command to specify the new docker registry, but Bob or John should comment on that. Bart From: Miller, Frank Sent: January 14, 2019 2:36 PM To: Qi, Mingyuan Cc: Friesen, Chris; Wensley, Barton; Church, Robert; Rowsell, Brent; Xie, Cindy; Penney, Don Subject: RE: Approach for SB 2004710 support to access docker images via proxy? Thanks for the update Mingyuan. Bart are you able to provide your suggestions to Mingyuan on the best approach for this? Frank From: Qi, Mingyuan [mailto:mingyuan.qi@intel.com] Sent: Monday, January 14, 2019 2:38 AM To: Miller, Frank Cc: Friesen, Chris; Wensley, Barton; Church, Robert; Rowsell, Brent; Xie, Cindy Subject: RE: Approach for SB 2004710 support to access docker images via proxy? Frank, Thanks for your note, I was just wondering how wide I should send my I thoughts for review. Last week, I tried 2 approaches to implement proxy: First one is following the complete mechanism of config_controller • collect proxy info in input_config() and treat them as host values. • add 3 attributes(http_proxy, https_proxy, no_proxy) to ihost of sysinv/cgts-client. • accordingly add 3 fields in sysinv host api/object/db. • add 3 params to docker puppet class and update them during pupput platform yaml file generation. • apply docker.pp to create http-proxy.conf with user input proxy info. One questions about this approach: Is ihost the suitable table for proxy info? Most likely the first controller is the only one node that docker needs proxy to access internet on, is this info redundant for each host? Or an alternative one without adding proxy info to sysinv db is better? I did a trail but introduced a shortcut mechanism in config_controller: • collect proxy info in input_config() as well • a utility script to add proxy info to host yaml after host puppet config creation finished. • Same as the previous approach to apply docker puppet manifest Same situation as the docker registry mirror, they could be done as the same approach. Really appreciate your comments. Thanks, Mingyuan From: Miller, Frank [mailto:Frank.Miller@windriver.com] Sent: Monday, January 14, 2019 6:36 To: Qi, Mingyuan <mingyuan.qi(a)intel.com<mailto:mingyuan.qi@intel.com>> Cc: Friesen, Chris <Chris.Friesen(a)windriver.com<mailto:Chris.Friesen@windriver.com>>; Wensley, Barton <Barton.Wensley(a)windriver.com<mailto:Barton.Wensley@windriver.com>>; Church, Robert <Robert.Church(a)windriver.com<mailto:Robert.Church@windriver.com>>; Rowsell, Brent <Brent.Rowsell(a)windriver.com<mailto:Brent.Rowsell@windriver.com>>; Xie, Cindy <cindy.xie(a)intel.com<mailto:cindy.xie@intel.com>> Subject: Approach for SB 2004710 support to access docker images via proxy? Mingyuan: Thank-you for taking this on. Can you describe the code changes that you think are needed for adding support to access docker images via a proxy? I’ve cc’d a few senior designers who can help you if required and provide initial feedback to you before you get too far into any implementation and post a gerrit review. Frank

2 2

[Starlingx-discuss] how to create feature branch remotely on StarlingX code repository
by Liu, Changcheng 15 Jan '19

15 Jan '19

Hi all, I need create remote feature branch for cooperation to upgrade Ceph. Does anyone know how to create feature branch remotely on StarlingX? B.R. Changcheng

3 4

[Starlingx-discuss] Sanity Test - ISO 20190115
by Alonso, Juan Carlos 15 Jan '19

15 Jan '19

Status of the Sanity Test for last CENGN ISO: bootimage.iso from 2019-Jan-15 (link<http://mirror.starlingx.cengn.ca/mirror/starlingx/master/centos/20190115T06…>) Sanity Test is executed in a Virtual Environment Status: GREEN Simplex Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 25 TCs [PASS] TOTAL: [ 30 TCs PASS ] Duplex Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 26 TCs [PASS] TOTAL: [ 31 TCs PASS ] Multinode Controller Storage Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 26 TCs [PASS] TOTAL: [ 31 TCs PASS ] Multinode Dedicated Storage Setup 04 TCs [PASS] Provisioning 01 TCs [PASS] Sanity 26 TCs [PASS] TOTAL: [ 31 TCs PASS ] ------------------------------------------------------------------ Regards. Juan Carlos Alonso

1 0

[Starlingx-discuss] Issues with build instructions
by Victor Rodriguez 15 Jan '19

15 Jan '19

Hi team I am following the image build instructions from : https://docs.starlingx.io/developer_guide/ If these are not the correct instructions please let me know. I am stuck at the point of: $ docker build --tag $USER:centos-mirror-repository --file Dockerfile . I am behind a proxy and i already set up docker for proxy ( i tested docker run hello-world and works fine ) and set up the Dockerfile as suggested: ENV http_proxy " http://your.actual_http_proxy.com:your_port " ENV https_proxy " https://your.actual_https_proxy.com:your_port " ENV ftp_proxy " http://your.actual_ftp_proxy.com:your_port " RUN echo " proxy=http://your-proxy.com:port " >> /etc/yum.conf However, I am still facing this error Cannot find a valid baseurl for repo: base/7/x86_64 Could not retrieve mirrorlist http://mirrorlist.centos.org/?release=7&arch=x86_64&repo=os&infra=container error was 12: Timeout on http://mirrorlist.centos.org/?release=7&arch=x86_64&repo=os&infra=container: (28, 'Resolving timed out after 30541 milliseconds') The command '/bin/sh -c yum install -y epel-release sudo vim-enhanced net-tools git /usr/bin/yumdownloader rpm-build rpm-sign deltarpm wget bind bind-utils && rm /etc/yum.repos.d/CentOS-Sources.repo /etc/yum.repos.d/epel.repo' returned a non-zero code: 1 Full log at : https://hastebin.com/omuzimenid.sql Any help is more than welcome Regards Victor Rodriguez

3 5

[Starlingx-discuss] Unable to create tenant network in case of a flat provider network
by Giuseppe Sannino 15 Jan '19

15 Jan '19

1 0