basic machine setup cookbook
vagrant plugin install vagrant-chef-zero
vagrant plugin install vagrant-omnibus
virtualbox
vagrant version is
$ vagrant -v
Vagrant 1.8.5
vagrant plugins version are
vagrant-chef-zero (2.0.0)
vagrant-omnibus (1.4.1)
berks vendor cookbooks
vagrant up --provision
Sometimes it fails some reason(mostly network problem I think).
vagrant provision
I'll hope to fix it and robust to finish the cookbooks.
apt-get -q -y --force-yes -o Dpkg::Options::='--force-confold' -o Dpkg::Options::='--force-all' install docker-engine=1.12.3-0~trusty
When I executed vagrant up , it tooks 1.5 hour. It does not need to any interaction. Just type command and wait.
I think most of time takes to download compressed(gzip) sailfish index (11G) and extract it.
I test it with the VM which is assigned 4 CPU and 8GB RAM. I write these assignment to Vagrantfile.
If you have enough resource, you can choose the method of create sailfish index via sailfish command.
Answer. Maybe machine resource is not enought to execute.
Almost case is that RAM is not enough.
HISAT2 require some times 32GB.
or
You try to run some jobs parallel.
vagrant halt
VBoxManage sharedfolder remove bayesvm_1.3.0 --name vagrant
VBoxManage export bayesvm_1.3.0 -o bayesvm_1.3.0-dev.ova
- Galaxy run in docker
- Galaxy run with systemctl
- Galaxy starts on boot by default setting
- Galaxy using port 28080, 29002
Galaxy is available port 28080.
http://localhost:28080 or http://yourbayeslinuxhost:28080
- Check port 80 is used or not.
- If port 80 is used, stop the service.
- Change port 28080 to 80 by editing
/usr/local/galaxy-bitwf/scripts/start_bitwf.sh.-p 28080:80to-p 80:80
To avoid conflict with other application.
Port 80 is using by httpd server (Apache httpd , Nginx is famous )
sudo /etc/init.d/docker-galaxy start
sudo /etc/init.d/docker-galaxy stop
- service
/lib/systemd/system/docker-galaxy.service
- Start script
/usr/local/galaxy-bitwf/scripts/start_bitwf.sh
- Stop script
/usr/local/galaxy-bitwf/scripts/stop_bitwf.sh
Run Galaxy following directory
/usr/local/galaxy-bitwf/scripts
| In host | In docker container | File or Directory | Note |
|---|---|---|---|
| /usr/local/galaxy-bitwf/scripts | /usr/local/galaxy-bitwf/scripts | directory | for share |
| /usr/local/galaxy-bitwf/scripts/2790.diff | /galaxy-central/2790.diff | file | Galaxy pull 2790 diff |
| /usr/local/galaxy-bitwf/scripts/job_conf.xml | /etc/galaxy/job_conf.xml | file | Job setting |
| /usr/local/galaxy-bitwf/export | /export | directory | for share |
| /usr/local/galaxy-bitwf/data | /data | directory | for share Readonly |
| /usr/local/galaxy-bitwf/scripts/setup_inside_container.sh | /galaxy-central/setup_inside_container.sh | file | Startup script |
-
Detail templates/start_bitwf.sh.erb
-
More configurability for containerized jobs by jmchilton · Pull Request #2790 · galaxyproject/galaxy
If you using NFS for data persistent, you need to care about order of NFS mount and docker galaxy start.
NFS needs to mount before docker galaxy start.
This case /data is on NFS
docker exec galaxy_bitwf_1.3.0 ls /data
If there is no data , it fail to start. To resolve problem
sudo /etc/init.d/docker-galaxy stop
docker rm galaxy_bitwf_1.3.0
sudo /etc/init.d/docker-galaxy start
If you want to start docker-galaxy , you need to disable automatic start on boot.
sysv-rc-conf docker-galaxy off
Create conda environment and install bwa
conda create -n myenvironment bwa=0.7.12
source activate myenvironment
bwa
conda info -e
conda config --get channels
conda config --get channels --system
conda remove -n myenvironment bwa
conda env remove -n myenvironment
| VirtualMachine | Cookbook Name | Cookbook Version |
|---|---|---|
| 1.2.0 | basicsetup | 1.2.0 |
| 1.3.0 | basicsetup | 1.3.0 |
| 1.4.0 | basicsetup | 1.3.0 and upgrade by hand |
Next Release 1.5.0 cookbook name change is scheduled.
see TODO
Copyright (c) [2015-2017] [Manabu ISHII] and RIKEN Bioinformatics Research Unit Released under the [Apache License Version 2.0]