[logback-user] SMTPAppender with multiple filters

[ceki]

Filters attached to an appender determine whether the containing 
appender sees events. For example, the following configuration snipped 
will cause appender A to only see events of level WARN.

  <appender name="A" class="...."/>
    <filter class="ch.qos.logback.classic.filter.LevelFilter">
      <level>WARN</level>
      <onMatch>NEUTRAL</onMatch>
      <onMismatch>DENY</onMismatch>
    </filter>
   </appender>

Note that ch.qos.logback.classic.filter.LevelFilter is an equality 
filter checking on Level.WARN. Appender A will never see events of level 
TRACE, DEBUG, INFO and ERROR.

Triggering of emails is determined by the evaluator [1] assigned to 
SMTPAppender. For an event to trigger a message, it must pass through 
all the filters attached to SMTPAppender and have the evaluator return true.

HTH,

[1] http://logback.qos.ch/manual/appenders.html#smtpAppender_Evaluator

On 12.04.2013 16:47, romeara wrote:
> Hi there,
>
> I'm having an issue with using multiple filters as you had outlined in a
> previous post:
>
> <appender name="SQL_TRANSACTION_DELAY_EMAIL"
> class="ch.qos.logback.classic.net.SMTPAppender">
>           <filter class="ch.qos.logback.classic.filter.LevelFilter">
>               <level>WARN</level>
>               <onMatch>NEUTRAL</onMatch>
>               <onMismatch>DENY</onMismatch>
>           </filter>
>           <filter
> class="com.annadaletech.taxfree.kioskapp.util.SQLTransactionDelayFilter">
>               <transactionDelayText>running time
> (millis)</transactionDelayText>
>           </filter>
>           ...
>       </appender>
>
> The first filter is looking for any log entry marked as [WARN ] and the
> second checks this message for my own user defined value and whether or not
> the email needs to be sent. However, though the second filter code is being
> accessed, the email is not passed from an ACCEPT filterReply.
>
> I've also read somewhere that apparently the SMTPAppender wont send an email
> unless its catching an ERROR message. If so, how can I alter this?
>
> Your thoughts appreciated,
>
>
> Ray.
>
>


-- 
Ceki
65% of statistics are made up on the spot