The default way of setting them is by creating file in WEB-INF folder (next to wro.xml). The way wro4j is configured is customizable, but this is out of the scope of this page.

The file is loaded during the filter initialization and is reused for all requests during application lifetime.

Table of Available Configuration Options

The following parameters are available:

Property Name Default Value Description
debug true boolean flag (former known as configuration), with possible values: true (DEVELOPMENT) or false (PRODUCTION). Find out more about differences at the bottom of this page
minimizeEnabled true Flag for turning minimization on/off.
gzipResources true accepted values are: true or false (case insensitive). When this flag is enabled response will be gziped.
resourceWatcherUpdatePeriod 0 integer value for specifying how often (in seconds) the resource changes should be checked. When this value is 0, the cache is never refreshed. When a resource change is detected, the cached group containing changed resource will be invalidated. This is useful during development, when resources are changed often. (since 1.4.8)
resourceWatcherAsync false A boolean which enables/disables asynchronous resource watcher. The true value does make sense when resourceWatcherUpdatePeriod is greater than 0. (since 1.7.3)
cacheUpdatePeriod 0 integer value for specifying how often (in seconds) the cache should be refreshed. When this value is 0, the cache is never refreshed.
modelUpdatePeriod 0 integer value for specifying how often (in seconds) the model (wro.xml) should be refreshed. When this value is 0, the model is never refreshed.
header computed by wro4j allow explicit configuration of headers (for controlling expiration date, etc). The implementation was inspired from [ here]. The headers can be defined using this format: <HEADER_NAME1>: <VALUE1> | <HEADER_NAME2>: <VALUE2> Example: Expires: Thu, 15 Apr 2020 20:00:00 GMT | cache-control: public
disableCache false DEPRECATED and removed since 1.7.6. Used only in DEVELOPMENT mode and allows you to disable the cache, this way any request will force the processing of the model and resources.
parallelPreprocessing false A flag for enabling parallel execution of pre processors which may improve overall performance, especially when there are slow preProcessors
connectionTimeout 2000 Timeout (milliseconds) of the url connection for external resources. This is used to ensure that locator doesn't spend too much time on slow end-point.(since 1.4.5)
managerFactoryClassName N/A Fully qualified class name of the {@link WroManagerFactory} implementation. When this value is not specified a default instance is used (BaseWroManagerFactory).
encoding UTF-8 Encoding to use when reading and writing bytes from/to stream
ignoreMissingResources true When this flag is disabled (false), any missing resource will cause an exception. This is useful to easy identify invalid resources.
ignoreEmptyGroup true When a group is empty and this flag is false, the processing will fail. This is useful for runtime solution to allow filter chaining when there is nothing to process for a given request. (since 1.4.5)
ignoreFailingProcessor false Available since 1.4.7. When this flag is true, any failure during processing will leave the content unchanged.
cacheGzippedContent false When this flag is enabled, the raw processed content will be gzipped only the first time and all subsequent requests will use the cached gzipped content. Otherwise, the gzip operation will be performed for each request. This flag allow to control the memory vs processing power trade-off. (since 1.4.4)
jmxEnabled true a flag used for turning on/off JMX.
mbeanName wro4j-<contextPath> The name of MBean object (how it is displayed in JMX console). If contextPath is empty, the name is wro4j-ROOT

Configuration options for ConfigurableWroManagerFactory

Property Name Default Value Description
preProcessors N/A A comma separated values describing pre processor aliases to be used during processing
postProcessors N/A A comma separated values describing post processor aliases to be used during processing
uriLocators servletContext,uri,classpath A comma separated values describing locators aliases to be used during processing
namingStrategy noOp The alias of the NamingStrategy used to rename bundles (for build time solution only) - available since 1.4.7
hashStrategy MD5 The alias of the HashStrategy used to compute ETags & cache keys - available since 1.4.7

What are the differences between DEVELOPMENT & DEPLOYMENT?

There are a couple of differences between those two modes:

* In DEVELOPMENT - you can add "?minimize=false" in request url in 
  order to turn off minimization on static resources, while in DEPLOYMENT you can't. 
* In DEVELOPMENT mode any runtime exception is logged and thrown further, while in DEPLOYMENT it is logged and the response is redirected to 404.