The SAPS Archiver component is responsible to collect data and metadata generated by tasks whose processing have either finished or failed. To this end, the Archive interacts with the SAPS Service Catalog, temporary and permanent storage pools.
In an apt-based Linux distro, type the below commands to install the Archiver dependencies.
sudo apt-get update
sudo apt-get -y install openjdk-8-jdk
sudo apt-get -y install maven
sudo apt-get -y install git
sudo apt -y install python-swiftclient
sudo apt-get install -y nfs-kernel-server
In addition to above Linux packages, download the own saps-engine
repository (which holds the Archive code). To fetch and compile the source code of it, follow the below steps:
git clone https://github.com/ufcg-lsd/saps-engine
cd saps-engine
git checkout develop
mvn install -Dmaven.test.skip=true
The first part of SAPS Archiver configuration deals with the setup of the NFS server temporary storage
. Before starting the NFS daemon, please choose an directory, $nfs_server_dir_path
, in your machine local file system. The NFS daemon will hold the exported files within this directory. Below commands setup the NFS server:
mkdir -p $nfs_server_dir_path
echo "$nfs_server_dir_path *(rw,insecure,no_subtree_check,async,no_root_squash)" >> /etc/exports
sudo service nfs-kernel-server restart
To check, list the directories mounted with nfs using the command below and check if the directory you just created appears.
showmount -e localhost
In the second part of the configuation, thearchiver.conf file should be customized to indicate how the Archiver should interact with other components, including the SAPS Catalog, temporary and permanent storage pools.
Once the configuration file is customized, the below commands are used to start and stop the Archiver component.
# Start command
bash bin/start-archiver
# Stop command
bash bin/stop-archiver