June 2021 - openSUSE Build Service - openSUSE Mailing Lists

[opensuse-buildservice] Add external source to a package?
by Carsten Ziepke 12 Sep '23

12 Sep '23

Hello, I want to update the package ja2-stracciatella in the games repository to version 0.17.0. https://build.opensuse.org/package/show/home:Kieltux:branches:games/ja2-str… The new version needs the library string_theory which is not available for openSUSE. The build process wants to download string_theory from github, because it is not found. [ 36s] Scanning dependencies of target string_theory [ 36s] [ 12%] Creating directories for 'string_theory' [ 37s] [ 25%] Performing download step (download, verify and extract) for 'string_theory' [ 37s] -- Downloading... [ 37s] dst='/home/abuild/rpmbuild/BUILD/ja2-stracciatella-0.17.0/build/dependencies/lib-string_theory/builder/string_theory-prefix/src/3.1.tar.gz' [ 37s] timeout='none' [ 37s] -- Using src='https://github.com/zrax/string_theory/archive/3.1.tar.gz' [ 37s] -- Retrying... Can I add string_theory (string_theory-3.1.tar.gz) manually to the package and patch the build process? Thank you, Carsten -- Well, since MS cant be sure of the username of someone downloading things, they are going to play it safe and have everything dowloaded and executed by Explorer as suid root. That way, it will run on ANY system anywhere. :) -- George Bonser <grep(a)cris.com>

3 4

Implementing a new APT publisher
by Arnaud Ferraris 30 Mar '22

30 Mar '22

Hi, We currently use OBS for building and publishing APT packages for the Apertis project[1] (a Debian derivative distribution). At the moment we use downstream patches for the publisher (src/backend/bs_publish) in order to use reprepro instead of dpkg-scanpackages/dpkg-scansources. As we're considering implementing a new APT publisher based on aptly[2], we'd like to do so in a way that can be upstreamed to OBS so it can benefit the whole community. This obviously raises several questions as to how we should proceed, so I'm sending this email as a call for advice/comments/suggestions. We're considering the following: - this new publisher would be implemented in addition to the existing dpkg-scan* based one and be selectable through global configuration options. - using aptly opens up possibilities, but it isn't always as straightforward as a few system() call; we would therefore need to either ship a helper tool or add a new BSPublisher submodule (the latter is the preferred option atm) - aptly can also be used through a REST API, which implies some asynchronous processing would be preferrable. Is there currently a way to handle async processing in OBS, or would that require another helper tool/module, if possible at all? We're only at the planning and design stage for now, so we would be very interested in your opinion and comments regarding the above. Thanks, Arnaud [1] https://apertis.org [2] https://aptly.info

3 3

x86_64 disk space errors
by Avindra G 11 Aug '21

11 Aug '21

I've been having trouble getting deno to build since v1.9.0 (April 13th, 2021) on x86_64. Tried various _constraints values, but the job is failing with the same error: [ 3311s] LLVM ERROR: IO failure on output stream: No space left on device [ 3311s] /.build/build-vm: line 460: echo: write error: No space left on device [ 3312s] error: could not compile `deno` [ 3312s] https://build.opensuse.org/package/show/devel:languages:javascript/deno#com… Can anyone advise? Thanks, avindra goolcharan

3 6

Metrics
by Moritz Röhrich 16 Jul '21

16 Jul '21

Dear OBS Devs, I was wondering if there is a way to collect metrics from OBS. Of particular interest to me are build queue length, number of in-progress package builds and regular CPU/memory usage metrics. In my use case, I want to use the metrics to let Kubernetes determine automatically how many workers to deploy. This works almost alright with the limited metrics support that I already have from Kubernetes itself, but there are some pitfalls and edge cases where Kubernetes removes workers before they are fully done building a package. My current situation is this: I have a StatefulSet for the workers and a HorizontalPodAutoscaler which will scale up the StatefulSet once CPU usage (as determined by Kubernetes itself) exceeds the threshold (workers are no longer idle). Due to long periods of IO bound operations where CPU usage may fall below the threshold, the autoscaler needs to delay scale-down operations by at least 30 min. This is quite hacky and only really works as long as the build time of the packages does not significantly exceed 30 minutes. The downsides are that the scaling only happens once CPU usage increases, which is usually quite late in the package building process and that workers may be killed too early, if CPU usage at the end of the build process is too low for too long. If I had the build queue length and the number of in-progress package builds as scrapeable metric, I could utilize that to let Kubernetes determine the right amount of workers more reliably. Best regards, Moritz Röhrich -- Quobyte GmbH, Berlin, AG Charlottenburg HRB 149012 B, Felix Hupfeld, Bjoern Kolbeck -- Quobyte GmbH, Berlin, AG Charlottenburg HRB 149012 B, Dr. Felix Hupfeld, Dr. Bjoern Kolbeck

3 4

SSL Fingerprint Mismatch when using DoD Repositories
by Rose, Robert 01 Jul '21

01 Jul '21

Hi all, Can anybody explain to me in how the SSL fingerprint check for DoD repositories works and why it is strongly recommended for security reasons in https://openbuildservice.org/help/manuals/obs-user-guide/cha.obs.concepts.h… We use DoD repositories to pin a tumbleweed snapshot we are working on in our private OBS instance. Each repository has the the respective Tumbleweed snapshot at https://download.opensuse.org/history/ together with the SSL fingerprint of download.opensuse.org<https://download.opensuse.org/history/> specified in its master subelement. However, since last week OBS refuses to download new DoD RPMs due to a fingerprint mismatch. dodup.log says: https://download.opensuse.org/history/20210622/tumbleweed/repo/oss/repodata…: peer fingerprint does not match: 82bac52cf77a1ec92d2184ebb574a97ba28ba98b7b63c3c67656a16ff708fca5 != 286869be29119fd068f94c0b1cd318068568c0b7f3036b303f5d102e512be80f Please notice that 82ba...fca5 is the (current!) fingerprint of download.opensuse.org<https://download.opensuse.org/history/> and 2868...e80f is from ftp.gwdg.de/<https://ftp.gwdg.de/>, which makes kind of perfect sense as a curl on https://download.opensuse.org/history/20210622/tumbleweed/repo/oss/repodata… on our download server gives the following response: ~> curl https://download.opensuse.org/history/20210622/tumbleweed/repo/oss/repodata… <!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN"> <html><head> <title>302 Found</title> </head><body> <h1>Found</h1> <p>The document has moved <a href="https://ftp.gwdg.de/pub/opensuse/history/20210622/tumbleweed/repo/oss/repod…">here</a>.</p> <hr> <address>Apache/2.4.43 (Linux/SUSE) Server at download.opensuse.org Port 443</address> </body></html> Now to my questions: Is OBS expected to check the fingerprint of documents that can be redirected? I would have expected it to check only the fingerprint of the repomd.xml file in the regarding repository. On other servers I get other redirects (e.g. http://ftp.halifax.rwth-aachen.de) Anyway, I fail to see the sense of an SSL fingerprint check when resources can redirect to other domains. Do I miss an important point here? Wouldn't it be sufficient to rely on the GPG signature of the RPMs? I am considering removing the master subelement completely. This issue does not have its origin in a rotated letsencrypt certificate, we already take that into account. However, a new certificate could be issued the upcoming days so please do not look to closely on the specific fingerprints mentioned in this post but the general questions about fingerprint checks in DoD repositories. Any help is greatly appreciated. Regards, Robert Appendix: Examplary DoD configuration in our OBS <repository name="next"> <download arch="x86_64" url="https://download.opensuse.org/history/20210623/tumbleweed/repo/oss/" repotype="rpmmd"> <master url="https://download.opensuse.org/history/20210623/tumbleweed/repo/oss/" sslfingerprint="sha256:82bac52cf77a1ec92d2184ebb574a97ba28ba98b7b63c3c67656a16ff708fca5"/> <archfilter>x86_64</archfilter> <pubkey> -----BEGIN PGP PUBLIC KEY BLOCK----- Version: GnuPG v2.0.15 (GNU/Linux) mQENBEkUTD8BCADWLy5d5IpJedHQQSXkC1VK/oAZlJEeBVpSZjMCn8LiHaI9Wq3G 3Vp6wvsP1b3kssJGzVFNctdXt5tjvOLxvrEfRJuGfqHTKILByqLzkeyWawbFNfSQ 93/8OunfSTXC1Sx3hgsNXQuOrNVKrDAQUqT620/jj94xNIg09bLSxsjN6EeTvyiO mtE9H1J03o9tY6meNL/gcQhxBvwuo205np0JojYBP0pOfN8l9hnIOLkA0yu4ZXig oKOVmf4iTjX4NImIWldT+UaWTO18NWcCrujtgHueytwYLBNV5N0oJIP2VYuLZfSD VYuPllv7c6O2UEOXJsdbQaVuzU1HLocDyipnABEBAAG0NG9wZW5TVVNFIFByb2pl Y3QgU2lnbmluZyBLZXkgPG9wZW5zdXNlQG9wZW5zdXNlLm9yZz6JATwEEwECACYC GwMGCwkIBwMCBBUCCAMEFgIDAQIeAQIXgAUCU2dN1AUJHR8ElQAKCRC4iy/UPb3C hGQrB/9teCZ3Nt8vHE0SC5NmYMAE1Spcjkzx6M4r4C70AVTMEQh/8BvgmwkKP/qI CWo2vC1hMXRgLg/TnTtFDq7kW+mHsCXmf5OLh2qOWCKi55Vitlf6bmH7n+h34Sha Ei8gAObSpZSF8BzPGl6v0QmEaGKM3O1oUbbB3Z8i6w21CTg7dbU5vGR8Yhi9rNtr hqrPS+q2yftjNbsODagaOUb85ESfQGx/LqoMePD+7MqGpAXjKMZqsEDP0TbxTwSk 4UKnF4zFCYHPLK3y/hSH5SEJwwPY11l6JGdC1Ue8Zzaj7f//axUs/hTC0UZaEE+a 5v4gbqOcigKaFs9Lc3Bj8b/lE10Y =i2TA -----END PGP PUBLIC KEY BLOCK----- </pubkey> </download> <arch>x86_64</arch> </repository> -- Robert Rose Embedded System Engineer Edge Device Competence Center

3 2

science repos
by Mark Olesen 30 Jun '21

30 Jun '21

Anyone know the reason that the science repo currently only goes up to SLE_SP3? Who decides when to add in SLE_SP4, SLE_SP5 etc? Cheers, /mark

2 2

ARM architectures not available for Ubuntu 18.04
by Kyle Edwards 29 Jun '21

29 Jun '21

Hello, I'm currently trying to build a package on Ubuntu 18.04 with x86 and ARM architectures (aarch64 and armv7l) enabled. The x86 builds work fine, but the ARM builds fail to start as "unresolvable" because none of the packages are available ("nothing provides <package>".) Are the ARM architectures not enabled on 18.04? I thought I'd heard that they were. Kyle

2 4

[2.11~alpha] can not create new packages - with solution (maybe)
by Mario Goppold 25 Jun '21

25 Jun '21

Dear OBS Devs, with obs-api-2.11~alpha.20210618T095403.f2091342ed-lp152.12530.1.noarch I can not create new packages in private instance. osc mkpac packageC cd packageC osc commit -m "initial in obs" Sending meta data... Server returned an error: HTTP Error 500: Internal Server Error Request: https://obs/source/home%3Amgoppold/packageC/_meta Headers: Date: Tue, 22 Jun 2021 17:57:32 GMT Server: Apache X-Request-Id: eea5880f-6f48-4bcf-8272-1a19a2d75d10 X-Runtime: 0.084162 X-Powered-By: Phusion Passenger(R) 6.0.8 Content-Length: 131 Status: 500 Internal Server Error Cache-Control: public Connection: close Content-Type: application/xml; charset=utf-8 Update / modifying existing packages works fine. The Error occur on cmdline and WebUI. production.log: I, [2021-06-22T18:14:57.924545 #2200] INFO -- : [d2b6d7f0-81dd-4e23-a4e6-042eb2f2944c] [2200:325.61] method=POST path=/package/create/home:mgoppold format=html controller=Webui::PackageController action=create status=500 error='ThinkingSphinx::SphinxError: index package_core: does not support DELETE (enabled=1) - DELETE FROM package_core WHERE sphinx_internal_id IN (1226)' duration=28.79 view=0.00 db=5.07 params={"authenticity_token"=>"[FILTERED]", "package"=>{"name"=>"packageC", "title"=>"", "description"=>""}, "commit"=>"Create", "project"=>"home:mgoppold"} host=192.168.156.10 time=522.79 backend=0 user=mgoppold *F, [2021-06-22T18:14:57.929467 #2200] FATAL -- : [d2b6d7f0-81dd-4e23-a4e6-042eb2f2944c] [2200:325.62] ** **[d2b6d7f0-81dd-4e23-a4e6-042eb2f2944c] ThinkingSphinx::SphinxError (index package_core: does not support DELETE (enabled=1) - DELETE FROM package_core WHERE sphinx_internal_id IN (1226)):* [d2b6d7f0-81dd-4e23-a4e6-042eb2f2944c] [d2b6d7f0-81dd-4e23-a4e6-042eb2f2944c] app/controllers/webui/package_controller.rb:295:in `create' [d2b6d7f0-81dd-4e23-a4e6-042eb2f2944c] config/initializers/wrap_parameters.rb:37:in `call' After the update, I also did the following: cd /srv/www/obs/api/ RAILS_ENV=production bin/rake migrate_options_yml RAILS_ENV=production bin/rails db:migrate:with_data Now I read the ReleaseNotes-2.11 ( README.SETUP is a broken link ) an do as suggested chroot --userspec=wwwrun:www / /bin/bash -c "cd /srv/www/obs/api/ && /usr/bin/bundle.ruby2.5 exec rails ts:rt:rebuild RAILS_ENV=production" The last one gives: using config file '/srv/www/obs/api/config/production.sphinx.conf'... listening on 127.0.0.1:9314 precaching index 'package_core' precaching index 'project_core' precached 2 indexes in 0.001 sec Started searchd successfully (pid: 3280). Generating index files for package_core rails aborted! ThinkingSphinx::ConnectionError: Error connecting to Sphinx via the MySQL protocol. Can't connect to MySQL server on '127.0.0.1' (115) bin/rails:4:in `require' bin/rails:4:in `<main>' Caused by: Mysql2::Error::ConnectionError: Can't connect to MySQL server on '127.0.0.1' (115) bin/rails:4:in `require' bin/rails:4:in `<main>' Caused by: Innertube::Pool::BadResource: Innertube::Pool::BadResource bin/rails:4:in `require' bin/rails:4:in `<main>' and the error on commit changes to: I, [2021-06-22T18:34:10.840529 #2200] INFO -- : [ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] [2200:1478.53] method=PUT path=/source/home:mgoppold/packageD/_meta format=xml controller=SourceProjectPackageMetaController action=update status=500 error='ThinkingSphinx::ConnectionError: Error connecting to Sphinx via the MySQL protocol. Can't connect to MySQL server on '127.0.0.1' (115)' duration=73.23 view=0.00 db=5.89 params={"project"=>"home:mgoppold", "package"=>"packageD"} host=172.17.6.27 time=1675.66 backend=0 user=mgoppold *F, [2021-06-22T18:34:10.861620 #2200] FATAL -- : [ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] [2200:1478.55] ** **[ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] ThinkingSphinx::ConnectionError (Error connecting to Sphinx via the MySQL protocol. Can't connect to MySQL server on '127.0.0.1' (115)):* [ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] [ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] app/models/package.rb:712:in `block in update_from_xml' [ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] app/models/package.rb:678:in `update_from_xml' [ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] app/controllers/source_project_package_meta_controller.rb:55:in `update' [ecaffcec-eb7a-443f-a4fa-3cd2e1bfc881] config/initializers/wrap_parameters.rb:37:in `call' That got me in the right direction: configure searchd (sphinx). Because my search "open build service configure searchd sphinx" didn't come up with anything useful I did: cat > /etc/systemd/system/searchd.service.d/override.conf <<EOF[Service] User=wwwrun Group=www EOF ln -sf /srv/www/obs/api/config/production.sphinx.conf /etc/sphinx/sphinx.conf systemctl enable --now searchd.service chroot --userspec=wwwrun:www / /bin/bash -c "cd /srv/www/obs/api/ && /usr/bin/bundle.ruby2.5 exec rails ts:rt:rebuild RAILS_ENV=production" Which solved my problem. Was that all I had to do or should I set up some more? Shouldn't that be in a README or ReleaseNotes-2.11? Best regards, Mario P.S. Thank you for the nice pice of software!

2 2

Determining proper system constraints
by Mark Olesen 21 Jun '21

21 Jun '21

Is there a reliable means to estimate the system constraints necessary? I am currently build with some deb packs with these _constraints: <constraints> <hardware> <disk> <size unit="G">10</size> </disk> <memory> <size unit="G">8</size> </memory> </hardware> </constraints> For some debian/ubuntu variants it is fine, on others (not always reproducible) I get this: [17832s] /.build/build-recipe: line 204: echo: write error: No space left on device Could simply bump the constraints upwards (again), but it would be nice to eliminate the guesswork. Is there a proper means of estimating the requirements? Cheers, /mark

3 3

Broken repos after outage
by James Hodgkinson 18 Jun '21

18 Jun '21

I keep getting this error in my packages after the recent service outage, I can't seem to work out how to fix it: > service daemon error: > 400 remote error: /usr/lib/obs/service//tar.service No such file or directory (http://obs-service:5153/sourceupdate/home:yaleman/kanidm?timeout=7200) My repo's here: https://build.opensuse.org/package/show/home:yaleman/kanidm <https://build.opensuse.org/package/show/home:yaleman/kanidm> Under source files I also see: > Source Files > Sources could not be expanded: service error: 400 remote error: /usr/lib/obs/service//tar.service No such file or directory (http://obs-service:5153/sourceupdate/home:yaleman/kanidm?timeout=7200) If someone could please provide steps to troubleshoot/fix it, that'd be lovely. Thanks, James

3 3