openfoam-cloud 1.0

从Github安装

克隆存储库

git clone https://github.com/mikelangelo-project/openfoam-cloud.git

安装所需的python软件包

pip install -r requirements.txt

通过PIP安装

pip install openfoam-cloud

设置

安装以下必需的软件包

• capstan
• snap
• 格拉法纳：sudo apt-get install grafana
• 流入量：sudo apt-get install influxdb

复制本地设置

cp ofcloud/ofcloud/local_settings.py{.example,}

编辑本地设置以反映您的环境。你看到的地方 os.environ['env_variable_name']必须导出所需的值 为了你的环境。

OPENFOAM_BASENAME       = 'http://openfoam.example.com'
GRAFANA_BASENAME        = 'http://grafana.example.com'

S3_ACCESS_KEY_ID        = os.environ['S3_ACCESS_KEY_ID']
S3_SECRET_ACCESS_KEY    = os.environ['S3_SECRET_ACCESS_KEY']
S3_HOST                 = 's3-host'
S3_PORT                 = 8443

SNAP_SERVICE            = 'snap-api-root'
INFLUX_DB_HOST          = 'influxdb-host'
INFLUX_DB_PORT          = 8086
INFLUX_DB_NAME          = 'snap'
INFLUX_DB_USER          = 'root'
INFLUX_DB_PASS          = 'root'

# Dedicated openfoam network settings
OPENFOAM_NETWORK_PREFIX = 'openfoam'
OPENFOAM_NETWORK_CIDR = 'your desired network CIDR'
OPENFOAM_NETWORK_ALLOCATION_POOL_START = 'your IP allocation range start'
OPENFOAM_NETWORK_ALLOCATION_POOL_END = 'your IP allocation range end'
OPENFOAM_NETWORK_GATEWAY_IP = 'your gateway IP usually the first IP in CIDR range'

# Scheduler daemon settings
SCHEDULER_REFRESH_INTERVAL_SECONDS = 'desired scheduler daemon refresh interval in seconds'

# OpenFOAM simulations save their results on a NFS server as is evident from the NFS_IP setting. The
# LOCAL_NFS_MOUNT_LOCATION setting tells the scheduler daemon where to prepare simulation case files, capstan package etc.
# This folder should have the NFS location mounted (example /mnt/OpenFOAM_results) except when the scheduler runs on
# the NFS machine itself, then this can point directly to the exported directory (for instance '/export/OpenFOAM_results/')
LOCAL_NFS_MOUNT_LOCATION = 'path to folder with mounted NFS location'

# Network file storage server ip address
NFS_IP = 'nfs server ip'

# Location on the NFS server where OpenFOAM case files and results will be saved
NFS_SERVER_MOUNT_FOLDER = 'location on the nfs server, where simulation case files and results are saved'

# Maximum number of launch retries of one instance. When this limit is reached, the simulation instance enters the
# 'FAILED' state
OPENFOAM_SIMULATION_MAX_RETRIES = 3

# Overrides the number of maximum NOVA vcpu's used. Example: openstack allows use of maximum 24 vcpus. Our setting
# allows only 12. The scheduler_deamon will run new simulations until 12 vcpu's are used on nova.
# If this setting value is higher than nova's max VCPU quota, the latter will be respected.
OPENFOAM_MAX_CPU_USAGE = 18

# Overrides the number of maximum NOVA instances used. Example: openstack allows use of 10 instances. Our setting
# allows only 5. The scheduler_deamon will run new simulations until 5 oinstances are used on nova.
# If this setting value is higher than nova's max instance quota, the latter will be respected.
OPENFOAM_MAX_INSTANCE_USAGE = 8

运行

确保snap、infloxdb和grafana正在运行。为定义数据源 格拉法纳的大量涌入。

• 快照：/opt/snap/bin/snapd -t 0 -a /opt/snap/plugin
• 流入量：/etc/init.d/influxdb start
• 格拉法纳：^{TT5}$

然后启动OpenFoam后端

python manage.py runserver 0.0.0.0:8008

下一步启动openfoam调度程序后台程序

python manage.py runscheduler

致谢

这个项目是在riaMIKELANGELO project（编号645402）中进行的，开始于 2015年1月，由欧盟委员会根据 H2020-ICT-07-2014：先进的云基础设施和服务 节目。