- Add 'Android Lab' for Android development.
- Add 'IoT Lab' for IoT development.
- Add fail2ban support: rsyslog + fail2ban
- Reduce the lab size
- Remove noVNC support if possible
- Integreate noVNC outselvies and install necessary packages
- Fix up open files limitation, up to 1000?: /proc/sys/fs/inotify/max_user_instances
- Allow vncserver reuse the unix account for login: multi users on one container to save resources? not safety?
- Load variables for container via environment variables, not mount the configs/ into it, require to modify tools/docker/start.
- Add support for Docker Toolbox (Busybox environment) Need chroot to a Busybox rootfs and make cloud lab work in it. ref: https://github.com/boot2docker/boot2docker/releases
- Add notify feature for LXDE: libnotify-bin notify-osd $ DISPLAY=:1 notify-send 'The system will be turned off in 5 minutes' --urgency=critical -t 0
- Update the password in 5 minutes $ tools/docker/cmd linux-0.11-lab 'sh -c "echo ffffff > ~/.vnc/passwdfile; pkill x11vnc"'
- Notify and update the system $ tools/docker/cmd linux-0.11-lab 'sh -c "DISPLAY=:1 notify-send "System will be turned off in 5 seconds" -t 0; sleep 5; echo ffffff > ~/.vnc/passwdfile; pkill x11vnc"'
- Add multiple security levels
- Verify the auto start logic, /etc/rc.local?
- Fix up tools/docker/publish, .vnc-token path
- Add guacamole suppport, another web vnc implementation http://guacamole.incubator.apache.org/doc/gug/
- Fix up cgroup swap support warning edit /etc/default/grub, and set GRUB_CMDLINE_LINUX to "cgroup_enable=memory swapaccount=1", then run "sudo update-grub"
- Add io limitation
- lep-lab fix up perf report, ref: https://stackoverflow.com/questions/38927895/how-do-you-get-debugging-symbols-working-in-linux-perf-tool-inside-docker-contai enable kernel with sched statistic information add qemu-system-ARCH for lepd running and kernel customization, or simply use lep-lab with linux-lab.
- Fix up ip conflicts issue
- Need to add a group deploy feature, which can allocate different subnet for different group, then, the container can restored their own ip without conflicts in another group. http://blog.csdn.net/gobitan/article/details/51104362
- Or maintain a ip table, record the used ip addr in this table, udpate it while the container is removed.
- Use pipework? https://github.com/lzufalcon/pipework
- Simply try with https://linuxconfig.org/how-to-retrieve-docker-container-s-internal-ip-address
- Change the noVNC token:ip mapping method to improve the performance, no need to genrate a map file, simply use a dynamic convert function should work better.