CI-users November 2015

ci-users@lists.centos.org

9 participants
9 discussions

artifacts from inside and outside the ci infra
by Karanbir Singh 06 Jan '16

06 Jan '16

hi, what was the final decision around how best to export artifacts so they are visible on the same url as internally inside of ci ? regards, -- Karanbir Singh +44-207-0999389 | http://www.karan.org/ | twitter.com/kbsingh GnuPG Key : http://www.karan.org/publickey.asc

4 4

Introducing python-cicoclient: A library and CLI to admin.ci.centos.org
by David Moreau Simard 19 Dec '15

19 Dec '15

Heya, I read the docs, the mailing lists and glanced over the various scripts out there and was convinced there had to be a better way than using curl to communicate with admin.ci.centos.org. Introducing python-cicoclient: - Github: https://github.com/dmsimard/python-cicoclient - PyPi: https://pypi.python.org/pypi/python-cicoclient/ - Docs: http://python-cicoclient.readthedocs.org/en/latest/ I broke down this package in three main components: - client.py: Abstracts the http stuff - wrapper.py: Uses client.py to wrap around the available API calls and provide a consistent user experience - shell.py/cli.py: Provide a CLI interface around wrapper.py It can be used as both a CLI client with the provided 'cico' binary but also as a library - you can easily import wrapper and roll your own thing. Feel free to use it, contribute and provide feedback ! Please note that as of right now, there is a problem with the "node get" command due to an upstream bug, documented here: https://bugs.centos.org/view.php?id=9748 Please do not use "node get" until this bug is fixed :) David Moreau Simard Senior Software Engineer | Openstack RDO dmsimard = [irc, github, twitter]

4 10

upcoming CentOS Linux 7 release
by Karanbir Singh 02 Dec '15

02 Dec '15

hi guys, RHEL 7.2 was released a few days back and we've been busy sorting through the code and working on the downstream CentOS Linux release ( its going to likely be released in November and will be called 7.1511 ). I'd like to make that content available on ci.centos.org ahead of public release. And there are a couple of ways to do this : 1) either we update the provisioning service, so all requests for a centos linux 7 machine will give you a 7.1511 build ( its 7.1503 at the moment ), with all the updates etc posted in sync that way. This also means that mirror.centos.org inside the ci infra will look DIFFERENT from mirror.centos.org the real users see, since it will be setup to look like the next release. 2) we make the content available in a new set of repos, called CentOS.next or something; and the ci users are able to optionally tweak their build scripts etc to include or not include these new changes. 3) we dont make this content available before its publicly released. of all those bits, I like (1) best; (2) only delays the content from hitting the ci infra by a week or 10 days. We dont usually take longer than that from when the builds are done to when the distro update is released. (3) just means that if there are changes, failures or breakages in the ci tested code, we only find out at the same time as the users - which isnt nice. thoughts ? -- Karanbir Singh +44-207-0999389 | http://www.karan.org/ | twitter.com/kbsingh GnuPG Key : http://www.karan.org/publickey.asc

3 4

Request to add an IRC channel to ci.centos.org jenkins
by Ben Kero 25 Nov '15

25 Nov '15

Hi there, I am doing some work building some CI for the RDO OpenStack Puppet Modules, and would like to enable IRC notifications for my builds. The channel I would like to add is #rdo-opm-ci on Freenode. -- Ben Kero RedHat, Senior Software Engineer, OPM-CI

2 1

RFE Ability to select specific chassis
by John Trowbridge 25 Nov '15

25 Nov '15

Howdy folks. How feasible do we think $subject is? For RDO-Manager jobs it is pretty much near zero chance of success on a machine with 16G of RAM. This is especially true for our HA jobs. I made a bit of a hack to our centosci provisioner in khaleesi[1] to throw out all hufty chassis nodes since they only have 16G RAM. This is slightly wasteful though, and would be better to be something that could be selected via the API. Thanks, John Trowbridge [1] https://github.com/redhat-openstack/khaleesi/commit/14d19c317c0a3fb6b551f4b…

2 3

Upcoming Feature: extendable sessions
by Karanbir Singh 24 Nov '15

24 Nov '15

hi, I'm going to be working on extending the admin.ci API. The intention is to add the following calls: /Session/get?key=<api key>&alias=<some text> : this will create a new session for the api key specified, but will not allocate any nodes to this session. The alias param is optional, if specified the session can be referenced with that alias ( in addition to / rather than the ssid ). /Session/done?key=<api key>&ssid=<ssid | alias for session > : this will tear down the session, including all Nodes allocated to the session. The ssid can be specified as either the UUID based ssid returned from the /Session/get call or the alias requested. In addition to these two new calls, the /Node/get call will get an optional ssid= param; calling /Node/get with an ssid will add the requested nodes to the existing session, rather than create a new session. The ssid can be either the uuid based ssid allocated at /Session/get or an alias to an existing session. The /Node/done call will also get an optional additional param in node= where you can specify the hostname of the machine to be dropped from an existing session, without terminating the entire session. With these four changes going in, it would be possible to setup a session, add nodes to that session as they are needed, drop hosts as required, and then tear down the session when done. The alias option could be used for various things, including accounting by job name, job type etc. Existing scripts that use the API will not need any changes, there will be no change to the existing calls. -- Karanbir Singh +44-207-0999389 | http://www.karan.org/ | twitter.com/kbsingh GnuPG Key : http://www.karan.org/publickey.asc

3 3

Network stability/reliability accessing external resources
by David Moreau Simard 16 Nov '15

16 Nov '15

Hi, I've heard that network connectivity for accessing resources outside of the ci.centos.org infrastructure can sometimes be unreliable. I haven't yet experienced this first hand as the jobs I'm constantly monitoring and am responsible for have not yet been migrated there. A lot of things can be put within the ci.centos.org infrastructure if need be but it's unrealistic to have *every* resource there. Is anyone aware of ongoing issues ? Should we report jobs that fail due to network errors ? A recent example is this job that failed due to DNS resolution errors [1] "Could not resolve host: trunk.rdoproject.org; Name or service not known". Thanks, [1]: https://ci.centos.org/view/rdo/job/rdo_manager-promote-build-delorean-rdo-l… David Moreau Simard Senior Software Engineer | Openstack RDO dmsimard = [irc, github, twitter]

3 3

Emergency ci.centos.org Maintenance 12-Nov-2015 01:00 UTC (20:00 EST)
by Brian Stinson 13 Nov '15

13 Nov '15

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi All, We will be rebooting Jenkins to apply an update that contains patches for multiple security vulnerabilities: https://wiki.jenkins-ci.org/display/SECURITY/Jenkins+Security+Advisory+2015… We will pause accepting new jobs 1 hour beforehand to let things settle. We do not expect a long outage for this update. As usual, we will post updates to this list and be available in #centos-devel on freenode for any questions. Cheers! - -- Brian Stinson CentOS CI Infrastructure Team -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iQIcBAEBAgAGBQJWQ2UKAAoJEIMGvNKzCweMMxcP/i9X0f6ETaLqlZ+/3sJ4vJ9f l4u7vwqFAj+FkrxLyvWNp/NLOh2EiYtXIM7oenQaI3Iy3xGKPgmxy7hf6ptpFeqZ ndh+vYCFr3fcGopzCsnk1qi4ENkAE+X7QRCZVve3FKHcfDmfz5HHMhkf7xMdcK62 //t91f5bcS49hBCKHF1v1hdrd8e5QzeGWYycKGK0N3NEubH4ruml5IgSF51ylYJM derKn45E+bn5iINBGxiT/OEIdl22ksvXSKOc2bzp+kkrzj/kqW67kGjlPIVSYQMF A1UJQNe8QkTtAvPtRalffJtmJU1ERfkAW4Nvm90hB0CUvDytEmp66Ks08FiSgAJa BrnhRe4rId3xPl95+yLzaD0IXR/lYOCAa/0FOAE1ynwm5YYvdak1yuSUL69MUGX8 VhDE9JhpUFKq0n8wiE5rAxAB51Q0Jk2c42mfPhMm/NISIpq4DHUNAH/TlNvB4ffg t/AhSEo49D858Eqe3/oGj3c/DOx+e1G073IOqyj7FXNFj/DQ/bhVk304j5wevTC7 j1bMca9l16M/FJlqBDejki5a3JVWvI9+j/F0Z3BJhYaaoNqaYqRoL1PnrHPIlB7e WEP9fhy8VpPQ/UKpBgiXxBkzSHr5wOAlz4XCuOczU2JllufLcDUKFZfIqIOvbY3z +d4iIjPD9DqqTLmh9gbx =2h3g -----END PGP SIGNATURE-----

1 2

Maintenance Period: Pausing jobs in 15 Minutes
by Brian Stinson 02 Nov '15

02 Nov '15

Hi Folks, We'll be pausing jobs starting at 2300UTC in preparation for our Maintenance period. Stay tuned here, and to #centos-devel for updates. Cheers! Brian

2 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

CI-users November 2015