I will deploy a k8s clusters on three pine64 A+ boards, all boards are installed opensuse tumbleweed, and one as k8s-master two as k8s-minions. The whole cluster will use TLS and BRAC security mechanism.

Setup an iscsi server on minnowboard/espressobin/odroid c2 based on opensuse tumbleweed, use two Nics to support multipath Setup an iscsi initiator on odroid c2 and dm

The the LUKS UUID in /etc/crypttab. In general the file system UUID is already used in /etc/fstab. So using the LUKS UUID in /etc/crypttab is a small step towards consistency and should make the system more robust (e.g. avoid bugs like bsc #1063249).

Build Stages

Travis-ci.org currently has nice beta features that I've wanted to try for a [long while(#https://travis-ci.org/foursixnine/openQA/builds/280652163)...

To be able to offer real Tumbleweed and a tested Leap for a Raspberry Pi3 the images for it need to be tested with openQA. Since

During my trip to and from SUSECon 2017 I had been working on a pinctrl driver for Actions Semi S500, based on a previous pinctrl driver of mine for Realtek RTD1295. Goal of this project is to complete the driver with pin definitions and muxes.

In os-autoinst and installation/bootloader_svirt.pm we currently we support only one-disk scenarios like SLES installation to disk and deployment from a disk in JeOS & CaaSP scenarios. To support RAID tests we need support of adding multiple disks to VMs. Targeting Xen and Hyper-V.

Try various container operating systems & kubernetes distributions and compare them to SUSE MicroOS and CaaSP, respectively.

In lieu of official hosting which was requested a year ago, go ahead and refactor approach to work on AWS S3. The previous approaches used either soft or hard links to avoid duplicating files that remain unchanged between snapshots, but S3 is not a file-system and does not support links. The closest comparable concept are object level redirects, but managing and creating those would be combersome, slow, and result in a higher bill. To avoid that an alternative that does not require object level redirects to simulate links, but that does not requiring duplicating files would be preferable. In addition the tool built to provide such snapshots on S3 should be easily deployable and setup to run automatically on AWS.

Planning to do some deep learning course sessions e.g. fastai , google ML crash course etc. Also try to understand common tools (tensorflow, jupyter notebook, numpy, pandas, pytorch) and practices e.g. Convoluation neutral nets, SGD used to solve learning problems. Aim is to get ready for kaggle competition (https://www.kaggle.com/competitions) eventually to test out learning and develop intuition around categories of learning problems.

About 3 years ago I wrote some Python and Kivy code for my Samsung Galaxy S4 in QPython. Unfortunately, QPython updated and broke its kivy compatibility. I've been waiting for them to fix it, but mostly have given up. So instead, just investigate how to turn the kivy code into an .apk for installation on an Android device! There are directions for this on the kivy website.