Mercurial > gemma
view example_conf.toml @ 4161:64cd18281c76
Improve performance of row level security policies
Using constraint_column_usage instead of key_column_usage makes
the query twice as fast. I did not explore why. Let's just take it.
Using 'EXISTS(... WHERE ... = value) is often more efficient than
value IN(...) since it allows the inner query to be executed only
up to the point where it turns out to return more than nothing with
filtering directly in place.
author | Tom Gottfried <tom@intevation.de> |
---|---|
date | Fri, 02 Aug 2019 17:14:13 +0200 |
parents | 317d176ef38c |
children | 209b10f7bb2c |
line wrap: on
line source
# ---------------------------------------------------------------------- # Example configuration for gemma back end # ---------------------------------------------------------------------- # ---------------------------------------------------------------------- # Basic Setup: # Host and port to listen on: host = "0.0.0.0" #port = 8000 # Where to find the data of the web client (SPA) to serve: web = "./web" # File to persist session data: sessions = "/tmp/gemma_session.data" # Duration until sessions expire if not renewed #session-timeout = "3h" # ---------------------------------------------------------------------- # Database: # DB connection configuration: db-host = "gemma-db" #db-name = "gemma" #db-port = 5432 #db-ssl = "prefer" db-user = "meta_login" db-password = "geo2Serv" # ---------------------------------------------------------------------- # Mail setup: # SMTP connection configuration: #mail-host = "localhost" #mail-password = "SECRET" #mail-port = 465 #mail-user = "gemma" # Client data to use: #mail-from = "noreplay@localhost" #mail-helo = "localhost" # ---------------------------------------------------------------------- # OGC services setup: # Connection to GeoServer instance for internal OGC services geoserver-url = "http://gemma-geoserver:8080/geoserver" #geoserver-user = "admin" #geoserver-password = "geoserver" # Clean GeoServer setup on startup # Persisting this is mainly useful in some dev setups. # Cleaning up the GeoServer should be done with # an explicit --geoserver-clean on the command line. #geoserver-clean = true # Proxy settings for external OGC services #proxy-key = "SECRET" #proxy-prefix = "http://localhost:8000" # # Server is known on the outside as: # external-url = "http://localhost:8000" # ---------------------------------------------------------------------- # CORS setup: #allowed-origins = "http://example.com" # ---------------------------------------------------------------------- # SOAP/XML-based importer setup: # soap-timeout = "3m" # Schema for "Testclient imports" # schema-dirs = "$PATH_TO_SCHEMATA" # published-config ="$PATH/pub-config.json"