Configuration service: anfisa.json

Configuration of the service is organized via the file anfisa.json The file is formed in extended JSON format, extension mean:

  • full line comments starting with '//' are allowed

  • macro definition mechanism supported

The following options are supported.

  • file-path-def, dictionary

The directory macro ${HOME} is predefined; here we define more macros, in particular the macro ${WORK}:

"file-path-def": {"WORK": "${HOME}/../a-setup"},

  • host, string

  • port, int

Host and port are used in the stand-alone configuration to setup the server on a specific port, ignored in the server (UWIGI container) mode:

"host": "0.0.0.0",
"port": 8190,

  • html-base, string

Used in the internal UI: its HTML pages need to know their location in terms of URL addresses, just to make correct reference to REST API or other pages of the internal UI, in case of server setup with NextGen UI should end in ‘/​app’:

"html-base": "/anfisa/",

  • html-title, string

Title prefix used in the pages of the Legacy UI:

"html-title": "Anfisa",`

  • mongo-db, string

The database in MongoDB used by the system:

"mongo-db": "Anfisa",

  • data-vault, string

The location of the vault directory:

"data-vault": "${WORK}/vault",

  • run-options, list

    Some additional option to configure Anfisa service. Currently out of use:

    "run-options": []

  • job-vault-check-period, int

    Period between checks of vault data consistence, in seconds:

    "job-vault-check-period": 30,

  • igv-dir, string

Location of file with configuration of access to BAM-files, used in IGV-links. See discussion here: IGV direct support.

"igv-dir": "${HOME}/igv.dir",

  • auto-drop-datasets, list of strings

The option determines dataset names that can be dropped automatically via request adm_drop_ds. There can be multiple patterns in list, and dataset name is good for drop if any of patterns appears in its name. Patterns should be correct Python regular expressions:

  "auto-drop-datasets": ["(?i)test"]

(Example shows the case when dataset name should contain substring "test" in ignorecase mode)

  • solutions, string

Additional solutions configuration, only gene panels (based on Symbol) are supported in the current version. Uncomment this configuration block to set additional panels in the instance of the application:

"--solutions": {
    "panels": {
        "Symbol": {
            "dir": "${WORK}/panels/Symbol"
        }
    }
}

  • export, dictionary

    Configuration of export functionality

  • work-dir, string

    The directory where the service stores exported files

"export": {
    "excel-template": "${WORK}/export/SEQaBOO_output_template_20190317.xlsx",`
    "work-dir": "${WORK}/export/work"
}

  • dir-files, list

    Setup of the mechanism of forwarding files as request results:

    "dir-files": [
    ["/ui/images", "${HOME}/int_ui/images"],
    ["/ui", "${HOME}/int_ui/files"],
    ["--/doc/", "${WORK}/doc/"],
    ["--/ui", "${WORK}/ui"],
    ["/excel", "${WORK}/export/work"]],

    Comments for instructions in example:

    • "/ui/images", "/ui":

      Requests for images and other sources, actual in stand-alone case. Should transfer the content of files located in the specific directory in $ANFISA_HOME; used in the internal UI in the stand-alone mode; in the server mode the same task is solved by configuration of the “main server”, Nginx or Apache

    • "--/ui":

      Requests for the source files when the anti-cache mechanism is on; in the server mode, to be used in the internal UI instead of the previous instruction; (in server setup drop two leading ‘-’ to make it working, and comment out the previous instruction)

    • "excel":

      This line sets the directory used to place the content of exported Excel files, supposing that they are going to be immediately downloaded by an external client

  • mirror-ui, list

    This instruction turns the anti-cache mechanism on; it consists of the paths to the source and target directories for mirroring (drop two leading ‘-’ to make it working):

    "--mirror-ui": ["${HOME}/int_ui/files", "${WORK}/ui"],

  • druid, dictionary

    See the section about Druid in installation documentation

  • sphinx-doc-sets, list

    Here is list of documentation sets available for the instance. Each set is defined as dictionary with properties:

    • id: short identifier of documentation sets

    • title: title of the sets

    • url: external URL to documentation set

    By default current documentation sets are already built and available on Intenet, so the option url should define correct link to this set.

    It is correct to comment this option, and then the documentation set will be built locally on start of the service. In this case the following additional options are required in the document set descriptor (in case of url option set they do not have effect):

    • source: path to source of documentation

    • build: path to building place

    • path: locally defined URL to the documentation set built

    Note In case of local building of documentation sets:

    • uncomment "/doc" instruction in dir-files

    • make sure if sphinx is installed on machine

    Example:

    "sphinx-doc-sets": [
   {
       "id": "dev",
       "title": "Anfisa Installation&Administration Documentation",
       "url": "https://foromeplatform.github.io/documentation/anfisa-dev.v0.7/",
       "source": "${HOME}/doc/dev/",
       "build": "${WORK}/doc/dev/",
       "path": "doc/dev/"
   },
   {
       "id": "user",
       "title": "Anfisa User Documentation",
       "url": "https://foromeplatform.github.io/documentation/anfisa-user.v0.7/",
       "source": "${HOME}/doc/user/",
       "build": "${WORK}/doc/user/",
       "path": "doc/user/"
   }
]

  • logging, dictionary

    Some standard Python way to configure the logging of a service. Please pay attention to one specific line of this stuff:

    filename line contains the configuration of the path to the logging directory:

    "logging": {
    ...
    "filename": "${WORK}/logs/anfisa.log"
}

  • doc-report-css, string

  • doc-pygments-css, string

    These two options are used to configure styling of documentation pages for datasets:

    "doc-report-css": "${HOME}/int_ui/files/report.css",
"doc-pygments-css": "${HOME}/int_ui/files/py_pygments.css",