This page helps you to configure Cloudgene and describes all parameters of the
Cloudgene uses an embedded H2 database to store meta data about jobs and users:
database: driver: h2 database: data/cloudgene user: cloudgene password: cloudgene
For production we recommend to use a MySQL database:
database: driver: mysql host: localhost port: 3306 user: cloudgene password: cloudgene
If no mail server is set, new registered users are activated immediately and no confirmation links are sent. This can be activated by defining a local or remote SMTP mail server:
mail: smtp: localhost port: 25 user: username password: password # the email address that Cloudgene uses to send emails name: email@example.com
Change the name of your service. If you plan to run Cloudgene on a sub-directory, then you can change the url-prefix.
# the name of your service [default: Cloudgene] name: My Service # max file size that can be uploaded in Bytes. -1 is unlimited [deafult: -1] uploadLimit: 50000 # root url of the web-application [default: empty] urlPrefix: /my-service
HTTPS Certificate and Security¶
Activate secure Cookies and use your own SSL Certificate to secure your connection:
# use https with the provided key store [default: false] https: true httpsKeystore: /your/key.jks httpsPassword: password # use secure cookies [default: false] secureCookie: true # use this secret key to generate JWT tokens. # please use a secret random string secretKey: some-random-string
More about on how to setup a java Keystore can be found here.
Directories and Workspace¶
If your service produces a lot of data, it could be useful to set the workspace directories to an other disc. The following directories can be changed:
# location for temporary files (e.g. cached file uploads) [default: tmp] tempPath: tmp # location for the results of a job [default: workspace] localWorkspace: /mnt/new-disc/workspace # HDFS location for the results of a job [default: cloudgene/data] hdfsWorkspace: cloudgene/data # if set all HDFS files are deleted after job execution [default: true] removeHdfsWorkspace: true # HDFS location for the meta-data of an app [default: cloudgene/apps] hdfsAppWorkspace: cloudgene/apps
Cloudgene manages two different queues to execute setup steps and workflow steps for a job. The number of jobs which are executed in parallel can be set for each queue independently:
# max. n jobs can execute their setup steps in parallel [default: 5] threadsSetupQueue: 5 # max. n jobs can execute their workflow steps in parallel [default: 5] threadsQueue: 5 # each user can run max. n jobs at the same time [default: 2] maxRunningJobsPerUser: 2
To change the default values please adapt the following parameters in your
# retire jobs after x days [default: 6] retireAfter: 6 # sent notification after x days [default: 4] notificationAfter: 4 # perform retire as a cronjob [default: false]. autoRetire: true # perform retire cronjobevery x hours [default: 5]. autoRetireInterval: 5
autoRetire is set to
false, you have to click on the Retire button in Administrator Dashboard to clean up.