CI-users February 2016

ci-users@lists.centos.org

12 participants
14 discussions

Strategy for scaling jenkins slaves ?
by David Moreau Simard 16 May '16

16 May '16

Hi, The RDO slave currently has 15 execution threads running on two cores and 4GB of RAM. As we use ci.centos more and more for gating and periodic jobs, I can see that the slave starts to struggle a little bit CPU-wise when the execution threads are busy. The bulk of the jobs are offloaded to an ephemeral bare metal node but ansible still runs from the slave. Some jobs do run locally, such as smaller tox jobs. What's the strategy if we want to scale this ? For example, do we want to scale in increments of "10 threads/2 cores/4 GB RAM" by adding more slaves as need be or do we want to grow the specifications of the existing slave and grow the amount of threads ? I have a personal preference to have more smaller slaves as this also has the added benefit of redundancy so we're not left entirely out in the cold if this one big slave happens to crash. David Moreau Simard Senior Software Engineer | Openstack RDO dmsimard = [irc, github, twitter]

5 7

Scheduled Maintenance 25-Feb-2016 08:00-10:00 UTC (03:00 - 05:00 Eastern)
by Brian Stinson 25 Feb '16

25 Feb '16

Hi Folks, We are taking ci.centos.org down for maintenance starting at 08h UTC on Thursday February 25th. We will pause jobs 1 hour beforehand to let the queue clear out. Planned work: - Update Jenkins to the latest LTS release and update all plugins to their latest versions. - Install the Github webhooks plugin for Jenkins - Add memory and reboot slave01 - Add memory and reboot the admin node (Duffy) We are hoping this will take less than the allotted 2 hours, but we're blocking out that time just in case. If your jenkins slave runs on slave01 we will restart the slave process for you, if you have another arrangement please be sure to double-check your process after this period. As always we'll hang out in #centos-devel to answer any questions or comments and we'll keep the list updated as things progress. Cheers! -- Brian Stinson CentOS CI Infrastructure Team

1 1

New f23 and f24 slaves for libguestfs
by Pino Toscano 24 Feb '16

24 Feb '16

Hi, I'd like to register new vm's/slaves for the libguestfs hypervisor: - fedora-23 (libguestfs-ci-f23slave) - fedora-24 (libguestfs-ci-f24slave) could you please provide credentials for them? Thanks, -- Pino Toscano

1 0

New version of python-cicoclient: 0.3.7
by David Moreau Simard 24 Feb '16

24 Feb '16

Hi, I've just drafted a new release for python-cicoclient, it is released on: - Github: https://github.com/dmsimard/python-cicoclient/releases/tag/0.3.7 - PyPi: https://pypi.python.org/pypi/python-cicoclient - COPR: https://copr.fedorainfracloud.org/coprs/dmsimard/python-cicoclient/ This new release provides support for retrying a "cico node get" in case of failures and documentation for that is available online at: http://python-cicoclient.readthedocs.org/en/latest/cli_usage.html#requestin… This feature is available both on the CLI and through the Ansible module. We will retry once by default in 10 seconds in case of failure. Thanks Dominic Cleal for the feature. David Moreau Simard Senior Software Engineer | Openstack RDO dmsimard = [irc, github, twitter]

1 0

duffy node/get and inventory APIs returning FQDN hostnames
by Vishwanath Bhat 24 Feb '16

24 Feb '16

Hi, Right now duffy apis like node/get and inventory gives us the hostnames and IP of the machines. But it gives short hostnames and not the resolvable FQDNs. It is a simple workaround to make it to FQDN. But still It would be good to have duffy return the resolvable FQDNs. Best Regards, Vishwanath

4 9

Question on testing in Vagrant with Jenkins
by karthik kumar 24 Feb '16

24 Feb '16

Hi all, My name is Karthik. In our company we are planning have our own t_functional testing running and add more tests and contribute back to the community. To start with planning to run test suite in Vagrant VM and got struck in pushing the test suite to Vagrant VM and run the ./runtest.sh I know I can have a "Execute Shell" block in Jenkins with vagrant ssh -c "git clone xxxx" and vagrant ssh -c "./runtests.sh" but it doesn't look the right way to do with Jenkins !! Isn't it ? My blind guess on whats being done in ci.centos is kickstarting machine/provisioning VM, making it as jenkins-slave and running the job on it ? So can you please advice how test suite is pushed to target Machine/VM ? Thanks

2 2

Provision to have an extra disk in the machine returned by duffy
by Vishwanath Bhat 24 Feb '16

24 Feb '16

Hi, I am using centos-ci infra for setting up CI for glusterfs (as part of Storage-SIG). And we need to have a proper disk for creating lvms which acts as gluster bricks. Right now duffy just gives you a machine with everything in root filesystem. This is not optimal for testing glusterfs for many reasons. Is it possible to request duffy to give a machine with small amount of storage for root and rest for our testing? It is Okay for us to wait for few minutes until duffy gives us the machines. Is there any workaround or any hack that we could do to get this done? Best Regards, Vishwanath

3 2

Nested virtualization in centos-ci?
by Honza Horak 17 Feb '16

17 Feb '16

Hi, I'm wondering if something like nested virtualization is possible/available/recommended in centos ci. My use case is that we would like to have a test for mariadb galera and this test is hard to be written on one machine + ideally we'd like to test real machine to machine workflow. Is it something we can do in the jenkins? Honza

4 4

Internal mirror for content you (ab)use within the CI env (aka adding cloud.centos.org)
by Fabian Arrotin 17 Feb '16

17 Feb '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In the last days, I had to grow the number of nodes behind cloud.centos.org in the centos.org infra, and when I added two more nodes, I saw that there were requests coming from the CI env to those cloud.centos.org nodes : "GET │ │/centos/7/devel/CentOS-7-x86_64-GenericCloud-GA-7.0.1406_01.qcow2 HTTP/1.1" 200 343368704 "-" "ansible-httpget" So, to speed up all those requests, I've created yesterday an internal mirror for cloud.centos.org, so that such requests will not go "outside" and use Internet bandwidth. If you have a need for an internal mirror for content you have to download for each of your tests, feel free to ask on this list, and we'll add it. That will mean faster download speed for your test[s] (internal to the DC) and also more internet bandwidth available for other tasks in that DC (so win/win) Kind Regards, - -- Fabian Arrotin The CentOS Project | http://www.centos.org gpg key: 56BEC54E | twitter: @arrfab -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iEYEARECAAYFAlbEGKAACgkQnVkHo1a+xU7AjQCdG/7pEM3GuUv5YOKFqmujRIkH 8I8AoIxdkbljxdFjG+eRXdIkHFLEXkWI =Uz+G -----END PGP SIGNATURE-----

1 0

Need few packages to be installed in centos jenkins slave
by Vishwanath Bhat 16 Feb '16

16 Feb '16

Hi, I am trying to run gluster tests from the Jenkins slave setup provided by centos infra. So in our case Jenkins slave also acts as a management node for the test framework. That node orchestrates the tests in each of the test machines. So in order to run the tests from the framework, I need few packages to be present in the slave machine and have passwordless ssh setup between the jenkins slave to all test machines. The second part it already taken care by the infra. We just need few packages to be installed on the slave. The slave is "slave01.ci.centos.org". And the packages required are plumbum, rpyc and unittest-xml-reporting. Since all of the above are python modules, I probably can use virtualenv and run the tests. I am going to try that tomorrow and update here. But in the meantime, if someone has access can install those packages, it would be of great help. Best Regards, Vishwanath

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

CI-users February 2016