ffRSAM is a web-based application to allow users to view plots of pre-calculated Real-time Seismic-Amplitude Measurement (RSAM) and frequency filtered (ff) RSAM. Two ways to view plots are provided:
Access to a Wave Server for data source is required.
Additionally, for the ffRSAM program, Docker and Docker Compose is required.
Podman and podman-compose will work as an alternative to Docker.
This application is meant to run on a centralized server so the website can be accessed by everyone in the observatory. It needs to run continuously to process new data so it is not suitable for desktop installation.
Minimum recommended computer specifications:
Select your installation directory (e.g. /opt is a good place for Linux) and clone ffrsam repository. Change directory into the respository.
cd /opt
git clone https://github.com/dnorgaard-usgs/ffrsam.git
cd ffrsamOr, if you have access to code.usgs.gov:
cd /opt
git clone https://code.usgs.gov/vsc/ffrsam.git
cd ffrsamCopy template.env to .env
cp template.env .envEdit the .env file.
Example (do not use the same passwords as below):
DB_USER=rsamuser
DB_PASSWORD=rsampassword
ROOT_PASSWORD=rootpwd
WEB_PORT=17000
DATA_DIR=./data
IMAGE_DIR=./images
IMAGE_WIDTH=1200
IMAGE_HEIGHT=400Notes:
Make sure the data and image directories specified exist. While this example places the data and images directory under /opt/ffrsam, you may want to consider creating these directories elsewhere on the server with more physical space.
cd /opt/ffrsam
mkdir data
mkdir imagesThe permissions on IMAGE_DIR specified must have writable permissions for all users. This is because the ffrsam-backend container that writes the images run as user with uid of 2020. Alternatively, you can create a user on your system with uid & gid of 2020 and set your image directory permissions writable by that user.
chmod a+w /opt/ffrsam/imagesor
groupadd -g 2020 ffrsam
useradd -u 2020 -g 2020 ffrsam
chown -R ffrsam:ffrsam /opt/ffrsam/imagesCreate config.py
cd backend/src
cp config_template.py config.pyEdit config.py
cd /opt/ffrsam
docker-compose build
docker-compose up -dVerify the 3 components are running:
D:\Programs\docker\ffrsam>docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
9ba90bbdd75d ffrsam_frontend "docker-php-entrypoi…" 39 minutes ago Up 39 minutes (healthy) 0.0.0.0:17000->80/tcp ffrsam-frontend
a1542ae979c1 ffrsam_backend "supercronic /app/cr…" 46 minutes ago Up 40 minutes ffrsam-backend
4d074b213ee7 mariadb:latest "docker-entrypoint.s…" 26 hours ago Up 40 minutes 3306/tcp ffrsam-dbThe following containers should be running:
The application can be accessed via the web port, e.g.:
http://localhost:17000 or http://hostname:17000
In order to ensure that other hosts on the network can access this, be sure to have the firewall open for port 17000 on the server.
For the pre-generated images, the day plot should begin populating after 10 minutes, the month plot after 1 hour, and year plot after 24 hours. Unless back-populating is performed (see below), images will show data only from day of installation forward.
On some systems the containers do not automatically start on reboot or restart of docker engine. It is best to configure each of the containers to start on boot.
For CentOS 7 create ffrsam-db.service, ffrsam-frontend.service, and ffrsam-backend.service files in /etc/systemd/system:
Example ffrsam-db.service:
[Unit]
Description = ffRSAM database
[Service]
WorkingDirectory = /opt/ffrsam/
ExecStart = docker start -a ffrsam-db
ExecStop = docker stop ffrsam-dbThe ffrsam-backend and ffrsam-frontend containers should ideally start after ffrsam-db container has started. Additional 'After' and 'Requires' statements may be helpful.
Example ffrsam-frontend.service:
[Unit]
Description = ffRSAM frontend
After = multi-user.target ffrsam-db.service
Requires = ffrsam-db.service
[Service]
WorkingDirectory = /opt/ffrsam/
ExecStart = docker start -a ffrsam-frontend
ExecStop = docker stop ffrsam-frontend
Example ffrsam-backend.service:
[Unit]
Description = ffRSAM backend
After = multi-user.target ffrsam-db.service
Requires = ffrsam-db.service
[Service]
WorkingDirectory = /opt/ffrsam/
ExecStart = docker start -a ffrsam-backend
ExecStop = docker stop ffrsam-backend
There are two ways to back populate ffRSAM data:
Connect to ffrsam-backend container.
docker exec -it ffrsam-backend /bin/bashRun ffrsam script from /opt/ffrsam.
Syntax: ./run.sh
cd /opt/ffrsam
./run.sh 1440 20200701000000 20200731000000In above example the first parameter, 1440 is the period in minutes, so 1 day period. This will also regenerate 1 year images. Allowed values for period:
The start and end time are simply in YYYYMMDDHHmmss format. Be sure to use a time range that is available in Winston.
To populate past or gap periods, you can use the backend/src/back_populate.* scripts. Edit the input parameters in back_populate.py:
Example:
# Input parameters
server=cfg.wws["AVO"]["server"]
port=cfg.wws["AVO"]["port"]
station='ACH'
channel='BHZ'
network='AV'
location='--'
start_time=UTCDateTime("20200626000000")
end_time=UTCDateTime("20200626040000") #YYYYMMDDHHMMSS
period=3600 # In SECONDSThen run either the back_populate.sh (Linux/Mac) or back_populate.bat (Windows) script. It will automatically kick off the back_populate.py in the ffrsam-backend container. It is recommended to run the file from a terminal or DOS prompt so that you can view the output as it runs. Check the output for any errors.
Note: Editing this file may prevent you from pulling future updates in git. If this happens simply revert the change with git checkout -- back_populate.py.
docker logs <container name>
docker logs -f <container name>docker exec -it <container name> /bin/bashOn CentOS 8, you can use Podman and podman-compose as an alternative:
yum install podman python
pip3 install podman-composeThen replace all "docker" commands with "podman" and "docker-compose" commands with "podman-compose".