devel

devel@lists.centos.org

12 participants
4610 discussions

Re: [CentOS-devel] [Pulp-list] [Discuss] Pulp/crane as container repo for CentOS Container Pipeline
by Mohammed Zeeshan 07 Mar '16

07 Mar '16

On Sun, Mar 6, 2016 at 10:13 PM, Michael Hrivnak <mhrivnak(a)redhat.com> wrote: > Mohammed, > > The problem is revealed in this strange-looking error, seen on line 16 of > your paste: "local error: no renegotiation". > > Docker is written in Go, the network library for which famously and > contentiously [0] does not support TLS renegotiation. > > So the challenge with docker is that you must host the API (crane in this > case) and the image files on web servers that will not require > renegotiation. > > [0] https://github.com/golang/go/issues/5742 > > Michael > > On Sun, Mar 6, 2016 at 4:55 AM, Mohammed Zeeshan < > mohammed.zee1000(a)gmail.com> wrote: > >> Hi, >> >> I am Mohammed Zeeshan Ahmed, part of Container Tooling Team at Redhat, >> working on the registry component of the CentOS Container Pipeline. >> >> I have been trying to get pulp/crane to work for sometime now to get it >> to fit into the pipeline. I have setup pulp/crane on a single node (self >> signed certificates) while a second node acts as a client (pulp-admin and >> docker) currently stuck at this point http://pastebin.centos.org/40926/ >> and would appreciate any help in getting things to work. >> >> Thanks in advance. >> >> -- >> *Mohammed Zeeshan Ahmed, * >> +919986458839 >> Bengaluru, India >> >> https://mohammedzee1000.wordpress.com/ >> <http://mohammed-zeeshan.strikingly.com> >> >> _______________________________________________ >> Pulp-list mailing list >> Pulp-list(a)redhat.com >> https://www.redhat.com/mailman/listinfo/pulp-list >> > > Hi, I have managed to resove the certificate issue, thanks for that help Michael, however, I think work still needs to be done. My current status, as things stand is http://pastebin.centos.org/40971/. Could use some more help to resolve this one. -- *Mohammed Zeeshan Ahmed, * B.E Computer Science Engineering Certified IT & Cloud Architect & RHCSA +919986458839 Bengaluru, India https://mohammedzee1000.wordpress.com/ <http://mohammed-zeeshan.strikingly.com>

2 3

Skylake install errors for CentOS-7
by Johnny Hughes 06 Mar '16

06 Mar '16

There were some install issues with some Intel Skylake chipsets/processors for CentOS-7. We have created a newer installer to fix this issue and the ISOs (CentOS-7-x86_64-DVD-1602-99.iso or CentOS-7-x86_64-Minimal-1602-99.iso) are available here: http://buildlogs.centos.org/centos/7/isos/x86_64/CentOS-7-x86_64-DVD-1602-9… http://buildlogs.centos.org/centos/7/isos/x86_64/CentOS-7-x86_64-Minimal-16… sha256sums: CentOS-7-x86_64-DVD-1602-99.iso: c226525cc5b47ffd32d8d082d8b316cb8f51592939f1b12d0dc846ff86d02e32 CentOS-7-x86_64-Minimal-1602-99.iso: 355eea6cc097e5f16805d075f27d2949aa17e8806127e4b361dcb2aef1eafe68 This is based on the latest released kernel (kernel-3.10.0-327.10.1.el7). This ISO set also has all current CentOS-7 updates rolled in including the new glibc. ========================== NOTE: Users who are not having issues what who want the latest ISOs should continue to use our normal ISO sets available here: http://buildlogs.centos.org/rolling/7/isos/x86_64/ Thanks, Johnny Hughes

4 3

inotify - triggering scripts based on file events
by michael gordon 06 Mar '16

06 Mar '16

We allow our third party clients to upload files for processing via ftp. Once the file is done a java app is triggered to process the file and move it to a processed folder. How do we know when the file is done writing,the file events we have observed include create, modify. These dont help since the ftp program writes file contents in blocks and then triggers the modified event, there is no finish event. What can you recommend -- *Michael Gordon* *Email: mgordonabc@ <gordonm(a)rada.gov.jm>gmail.com <http://gmail.com>*

3 2

[Discuss] Pulp/crane as container repo for CentOS Container Pipeline
by Mohammed Zeeshan 06 Mar '16

06 Mar '16

Hi, I am Mohammed Zeeshan Ahmed, part of Container Tooling Team at Redhat, working on the registry component of the CentOS Container Pipeline. I have been trying to get pulp/crane to work for sometime now to get it to fit into the pipeline. I have setup pulp/crane on a single node (self signed certificates) while a second node acts as a client (pulp-admin and docker) currently stuck at this point http://pastebin.centos.org/40926/ and would appreciate any help in getting things to work. Thanks in advance. -- *Mohammed Zeeshan Ahmed, * +919986458839 Bengaluru, India https://mohammedzee1000.wordpress.com/ <http://mohammed-zeeshan.strikingly.com>

1 0

How to build CentOS 7 AMI
by Alan Ivey 04 Mar '16

04 Mar '16

Hi everyone, I'd like to revisit the thread about how the CentOS 7 AMIs are created ( https://lists.centos.org/pipermail/centos-devel/2015-July/013652.html) and see if the process can be published in the https://github.com/CentOS/sig-cloud-instance-build repository or another relevant location. With CentOS 7 AMIs only being available in the Marketplace, all resulting EC2 instances have the Marketplace codes attached to the EBS volumes. A significant restriction of this is that a resulting image cannot be the non-primary volume of an instance unless it is powered down. This presents itself to be a problem in at least the following scenarios: - Unable to attach a CentOS 7 boot volume to another instance for repair without either creating a temporary instance or shutting down an existing one. For example, if you messed up the /etc/sudoers file and logged out, and wanted to repair, you would not be able to repair by mounting to another instance and editing the file without incurring additional (albeit small) cost, or having an existing instance be temporarily unavailable. - The "amazon-chroot" Packer Builder ( https://www.packer.io/docs/builders/amazon-chroot.html) does not work because it starts by mounting a copy of the snapshot tied to the AMI as part of a scripted operation and therefore cannot power off to do so Custom AMIs, snapshots, copied EBS volumes, etc, all have the marketplace codes copied to them and inherit the restrictions. If an org was to use these features for automating environments and was disconnected from the original Marketplace agreement, they may be unaware of this limitation. I would also appreciate being able to have the additional transparency of seeing how an AWS AMI is created as the docker/openstack/etc images from the repository referenced above. This would be useful in environments with regulatory compliance concerns, such as AWS GovCloud, HIPAA, FedRAMP, etc. I understand the benefit that Marketplace registrations allow for the ability to notify users of any changes, and I am not necessarily advocating for switching away from the Marketplace as the primary AMI location. I would like to be provided the opportunity to build a private AMI in the exact same procedure as the official image so as to avert the restrictions provided by the Marketplace. *[Note: I previously posted this question to centos-virt <https://lists.centos.org/pipermail/centos-virt/2016-February/004907.html>, but I did not receive any feedback and I have since noticed that **https://wiki.centos.org/SpecialInterestGroup/Cloud <https://wiki.centos.org/SpecialInterestGroup/Cloud> refers to the CentOS-Devel list for discussion. My apologies to subscribers of both lists.**]* Thank you, Alan

3 5

Re: [CentOS-devel] CentOS-devel Digest, Vol 133, Issue 4
by Brady Wied 04 Mar '16

04 Mar '16

Hi, Sorry for omitting the license. It's a BSD 2 clause and I've posted it here: https://github.com/wied03/centos-package-cron/blob/master/LICENSE.txt Hopefully that license is OK. -Brady On Fri, Mar 4, 2016 at 5:00 AM, <centos-devel-request(a)centos.org> wrote: > Send CentOS-devel mailing list submissions to > centos-devel(a)centos.org > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.centos.org/mailman/listinfo/centos-devel > or, via email, send a message with subject or body 'help' to > centos-devel-request(a)centos.org > > You can reach the person managing the list at > centos-devel-owner(a)centos.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of CentOS-devel digest..." > > > Today's Topics: > > 1. kickstart files for centos 7 cloud images (Matt Bacchi) > 2. RPM Submission - Centos Package Cron (Brady Wied) > 3. Re: RPM Submission - Centos Package Cron (Nux!) > 4. Re: RPM Submission - Centos Package Cron (Sven Kieske) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Thu, 3 Mar 2016 13:40:04 -0500 > From: Matt Bacchi <mbacchi(a)gmail.com> > To: centos-devel(a)centos.org > Subject: [CentOS-devel] kickstart files for centos 7 cloud images > Message-ID: > < > CANddqJcxy2jpToM06LxxH0jKzr9EAozwHJX9ku2Gey-Vg96htw(a)mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > Hello, > > I'm wondering where I can find the kickstart files used to create the cloud > images here: > > http://cloud.centos.org/centos/7/images/ > > I've created my own centos 7 image from kickstart but the instance doesn't > broadcast dhcp requests (verified using tcpdump.) If I launch an instance > using the image you provide (CentOS-7-x86_64-GenericCloud-1511.qcow2 > < > http://cloud.centos.org/centos/7/images/CentOS-7-x86_64-GenericCloud-1511.q… > >) > it does broadcast a dhcp request and then comes up on the network. My > kickstart file does indicate "network --bootproto=dhcp --onboot=on", but I > want to compare it to yours which is working as I would expect. > > Thanks in advance. > > -Matt >

1 0

kickstart files for centos 7 cloud images
by Matt Bacchi 03 Mar '16

03 Mar '16

Hello, I'm wondering where I can find the kickstart files used to create the cloud images here: http://cloud.centos.org/centos/7/images/ I've created my own centos 7 image from kickstart but the instance doesn't broadcast dhcp requests (verified using tcpdump.) If I launch an instance using the image you provide (CentOS-7-x86_64-GenericCloud-1511.qcow2 <http://cloud.centos.org/centos/7/images/CentOS-7-x86_64-GenericCloud-1511.q…>) it does broadcast a dhcp request and then comes up on the network. My kickstart file does indicate "network --bootproto=dhcp --onboot=on", but I want to compare it to yours which is working as I would expect. Thanks in advance. -Matt

1 0

AltArch ppc (32-bit) multilib support available
by James O'Connor 02 Mar '16

02 Mar '16

The ppc (32-bit) big endian power7 optimized rpms were signed and pushed to the ppc64 updates repo yesterday http://mirror.centos.org/altarch/7.2.1511/updates/ppc64/Packages/ We will find a place similar to buildlogs.centos.org to push all the ppc rpms, mock configs and build logs used in to the bootstrap process.

1 0

dm-cache not caching data on 3.18.21-17.el6.x86_64
by Thanos Makatos 02 Mar '16

02 Mar '16

I'm evaluating dm-cache on kernels 3.18.21-17.el6.x86_64 (Xen 4) and 2.6.32-573.7.1.el6.x86_64 (KVM). The test I do is a simple sequential read using dd(1). read_promote_adjustment and sequential_threshold have been set to 1 and 0, respectively. For the 2.6.32 kernel, all seems to be working fine, "#usedcacheblocks" correctly reflects the number of cached blocks based on what I'm reading with dd(1) performance is pretty much native SSD performance. However, the same test on the 3.18.21-17.el6.x86_64 kernel results in "#usedcacheblocks" being stuck to "2" and no performance improvement is observed. Any ideas what could be wrong? How can I further debug this? -- Thanos Makatos

1 0

[Discussion-CCCP] Pulp/crane story
by Mohammed Zeeshan 02 Mar '16

02 Mar '16

Hi, I just got off a half hour discussion with the pulp guys for the CentOS Container Pipeline registry story. 1. docker save x.tar produces a v1 tarball and not a v2 tarball 2. This tarball doesnt contain metadata such as tags and so on. 3. docker pull from docker 1.6 assumes the registry being pulled from is a v2 registry 4. v1 registry has been depricated. 5. Currently, the only way to get a v2 repo into pulp or any other registry is to first push it to docker registry and then sync our registry to it. I seriously am beginning to hate docker. -- *Mohammed Zeeshan Ahmed, * B.E Computer Science Engineering Certified IT & Cloud Architect & RHCSA +919986458839 Bengaluru, India https://mohammedzee1000.wordpress.com/ <http://mohammed-zeeshan.strikingly.com>

4 8

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

devel