- logback-user - mailman3.qos.ch

Changing the log level on a per-message basis
by Jason R Briggs 03 Dec '10

03 Dec '10

Hi Is there a reason why LoggingEvent.setLogLevel is a one-shot process? I'm writing a converter which should be able to modify the level of the log message (based on config), however that won't work because the method will throw an illegal arg exception on subsequent access. Or is there a better way to do this? Thx, Jason

1 0

Dynamically attaching the loggers to appenders
by Pradnya Gawade 02 Dec '10

02 Dec '10

Hi, I am having the issue described below, sorry for the long post but I think that is required to better explain the problem. I want to separate out the logs in such a way that some special packages, classes needs to go in specific log format and rest of the logs will go in default format. I have created the multiple file, Syslog appenders with all the required formats. But I am having difficulty configuring in such a way that the log messages from packages other than my special packages will go with the default appender. Initially my idea was to map the basic outer package to all the special appenders as well as to default appender, create a filter for each special appender so it will only take the specific package logs only. But the problem is that default appender takes all the log messages. I can again implement a filter for the default appender so it will take all the excluded log s only but probably it it will be lot of processing unnecessarily, evaluating the same criteria again and again for each logging request. Is there any way so I can intercept each logging request such that I can decide which appender I need this logger to map to (by implementing some java routine), put the result in MDC context and make use of that result variable in the configuration to associate the required appender (using if-else syntax)? I saw that Sifting appender might be helpful in this scenario but I don't know where I can put my processing logic to decide which appender I want the logger to associate to so I can use its result as discrminator. Any ideas? Thanks, Pradnya

1 1

Invitation to connect on LinkedIn
by vaibhav malhotra a través de LinkedIn 28 Nov '10

28 Nov '10

LinkedIn ------------vaibhav malhotra pidió añadirte como contacto en LinkedIn: ------------------------------------------ Ramon, I'd like to add you to my professional network on LinkedIn. - vaibhav Aceptar invitación de vaibhav malhotra http://www.linkedin.com/e/-fcbrfz-gh27ks5w-49/GD6y5Vns0zvc64fQGM60TpHfVbhLK… Ver invitación de vaibhav malhotra http://www.linkedin.com/e/-fcbrfz-gh27ks5w-49/GD6y5Vns0zvc64fQGM60TpHfVbhLK… ------------------------------------------ ¿SABÍAS que LinkedIn puede ayudarte a encontrar respuestas a tus preguntas más difíciles? Configurar tu perfil como público significa que tu perfil de LinkedIn aparece cuando la gente introduce tu nombre en los principales motores de búsqueda. ¡Controla tu imagen! http://www.linkedin.com/e/-fcbrfz-gh27ks5w-49/ewp/inv-22/ -- (c) 2010, LinkedIn Corporation

1 0

KnowledgeKnowledgeSession.startProcess() thread safe ?
by TheMax-2 26 Nov '10

26 Nov '10

Hello, I wrote a micro benchmark in order to test the performance of Drools Flow 5.1.0 in single and multi-threaded environments. And I noticed that, when 2 threads call startProcess() at the same time (Intel Core 2 Duo CPU), the returned process instance IDs are the same. I checked the code and it seems there is no locking mechanism in the org.drools.process.instance.impl.DefaultProcessInstanceManager class for the id creation. Is this a bug or am I missing something ? Thanks -- View this message in context: http://old.nabble.com/KnowledgeKnowledgeSession.startProcess%28%29-thread-s… Sent from the Logback User mailing list archive at Nabble.com.

1 0

Elongated system startup-time because of Logback configuration
by nikhil diwan 25 Nov '10

25 Nov '10

Hi, I have a startup_batch file in which I make a call to JAVA application which is responsible for starting up my application with SWT window which displays system running information. Issue is the elongated start up time for application, as I executes my batch file, it stops for a while (as if it's doing nothing) though I know that in background batch file is loooking for logback.grrovy or logback.xml file in the classpath then configures it, but meanwhile my SWT window is not even started and I need to wait for around 45-50 Secs whenever I restart my application. As I can't afford delay in application start so I am looking for some workaround to reduce this logback configuration time by my application everytime. Any helps/inputs/thoughts are welcome. ~Thanks Nikhil

6 9

Can the SiftingAppender discard 'unknown' log events?
by Colm.McDonnell＠rbs.com 24 Nov '10

24 Nov '10

I am using the SiftingAppender to isolate log events with a specific MDC key. This works really well, giving me high signal:noise ratio log files for specific invocations on particular process etc etc. However the SiftingAppender (or rather the MDCBasedDiscriminator) insists on the provision of a 'default value'. If I supply this default value (e.g. "unknown") then all unmatched events will be directed to unknown.log. But the 'other' log events are already logged separately so the entries in 'unknown.log' are unwanted duplicates (for me). Ideally I would use the sifting appender for a particular use case and let everything else be logged the way it is currently logged i.e. I would lke to provide no default value and have the appender understand this to mean 'ignore any events which do not match the given key' but that seems not to be supported so I'm trying to figure out how to either discard the unwanted events or to incorporate my existing logging into the sifting appender. I'd be grateful if someone could suggest how I might do either of the following: * somehow discard the 'unknown' events discriminated by the SiftingAppender i.e. to ignore them completely as they are already logged elsewhere or * figure out how to control the 'unknown' log events so that Ii can direct them to an 'everything else' appender which is distinct from the 'isolated events' appender I've included an example of the config below. The STDOUT appender is the main appender, there's a wrapper process running the java app and this wrapper process redirects STDOUT to its own log file (with its own rolling rules etc) therefore the 'everything else' events are all logged to STDOUT. I'd like the SIFTER to only log events for which it gets a match otherwise we end up with the same events in STDOUT and in "unknown.log". Alternatively, if there is a way to configure SIFTER such that it wraps the STDOUT appender and the unmatched events are only directed to that appender then that would work too. I have tried several configurations but I cannot associate the unmatched events with a specific appender. <configuration> <appender name="STDOUT" class="ch.qos.logback.core.ConsoleAppender"> <encoder> <pattern>[%-5p] [%t] %logger{5}: %m%n</pattern> </encoder> </appender> <appender name="SIFTER" class="ch.qos.logback.classic.sift.SiftingAppender"> <discriminator> <Key>isolatedLogName</Key> <defaultValue>unknown</defaultValue> </discriminator> <sift> <appender name="${isolatedLogName}" class="ch.qos.logback.core.FileAppender"> <File>${log.path:-target/log}/${isolatedLogName}.log</File> <encoder> <Pattern>%d | %-5p | %logger{5} | %m%n</Pattern> </encoder> </appender> </sift> </appender>  <root level="INFO"> <appender-ref ref="STDOUT" /> <appender-ref ref="SIFTER"/> </root> </configuration> For example: I tried this: <appender name="SIFTER" class="ch.qos.logback.classic.sift.SiftingAppender"> <discriminator> <Key>isolatedLogName</Key> <defaultValue>everythingElse</defaultValue> </discriminator> <sift> <appender name="isolatedLogName" class="ch.qos.logback.core.FileAppender"> <File>${log.path:-target/log}/${isolatedLogName}.log</File> <encoder> <Pattern>%d | %-5p | %logger{5} | %m%n</Pattern> </encoder> </appender> <appender name="everythingElse" class="ch.qos.logback.core.ConsoleAppender"> <encoder> <pattern>[%-5p] [%t] %logger{5}: %m%n</pattern> </encoder> </appender> </sift> </appender> And logback created two files - everythingElse.log and <the real value associated with the isolatedLogName key>.log - then wrote all output to the "everythingElse" appender. I was trying to tell it to do this: "if the key's value is the default value then ignore it or if that is not possible then direct the output to the appender named after that value otherwise direct to the other appender". Any help appreciated, thanks. *********************************************************************************** The Royal Bank of Scotland plc. Registered in Scotland No 90312. Registered Office: 36 St Andrew Square, Edinburgh EH2 2YB. Authorised and regulated by the Financial Services Authority. The Royal Bank of Scotland N.V. is authorised and regulated by the De Nederlandsche Bank and has its seat at Amsterdam, the Netherlands, and is registered in the Commercial Register under number 33002587. Registered Office: Gustav Mahlerlaan 10, Amsterdam, The Netherlands. The Royal Bank of Scotland N.V. and The Royal Bank of Scotland plc are authorised to act as agent for each other in certain jurisdictions. This e-mail message is confidential and for use by the addressee only. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer. Internet e-mails are not necessarily secure. The Royal Bank of Scotland plc and The Royal Bank of Scotland N.V. including its affiliates ("RBS group") does not accept responsibility for changes made to this message after it was sent. Whilst all reasonable care has been taken to avoid the transmission of viruses, it is the responsibility of the recipient to ensure that the onward transmission, opening or use of this message and any attachments will not adversely affect its systems or data. No responsibility is accepted by the RBS group in this regard and the recipient should carry out such virus and other checks as it considers appropriate. Visit our website at www.rbs.com ***********************************************************************************

3 4

Use web application name in logback configuration
by Pradnya Gawade 23 Nov '10

23 Nov '10

Hi, Can I use my Spring web application name in the logback configuration dynamically? I know logback can read from properties file specified under <property> or it can read from the system variables but is it possible to use some dynamic data like web application name in the configuration? Thanks, Pradnya

2 3

Delete log file
by pwillems66＠zonnet.nl 19 Nov '10

19 Nov '10

> It is not clear to me what you mean by "test". Is a test the class > containing the various test methods or does "test" refer to each test > method? For example, Given the MyTest class mention in my previous > post, is MyTest the "test" or are the methods a() and b() "tests"? > If the latter, you could try, > public class MyTest { > Logger logger = LoggerFactory.getLogger(MyTest.class); > > @Test a() { > MDC.put("testname", "myTest_a"); > logger.info("a says hello"); > ... > } > > @Test b() { > MDC.put("testname", "myTest_b"); > logger.info("b says hello"); > ... > } > } Let me try to explain what I am after: First: For me a "test" is a process. Executing 1,2,3,n-1,n method which could log information. But to keep it simple, executing method a() in class MyTest is a test (Create/append/truncate file myTest_a.log) An other test, executing method b() in class MyTest is a test (Create/append/truncate file myTest_b.log) Second: Lets use your class and methods for a simple application: Public Class MyApp { public static void main(String args[]) { MyTest myTest = new MyTest(); myTest.a(); myTest.b(); if (errors in method a) { delete file (myTest_a.log) myTest.a(); } } } If you run the code in file myTest_a.log there are 2 lines mentioned: a says hello a says hello because the file is in use. In case there are errors the (call myTest.a() again) the file myTest_a.log must be deleted and the next call must recreate the file. -Patrick

1 0

Problem with Syslog appender
by Pradnya Gawade 18 Nov '10

18 Nov '10

Hi, I have following requirements when sending the log from a java web application to syslog server: 1. When logging the stack trace in case of exceptions, all the stack strace should be seen as a single message. 2. Need to send either or all of the following information in logs- name of the web application, URL, version of the application. As per my knowledge, Syslog appender of Logback doesnot fit to above requirements. I am also looking for any other logging APIs which can satisfy these requirements. Any ideas? Thanks in advance. - Pradnya

3 7

Re: [logback-user] Delete log file
by pwillems66＠zonnet.nl 18 Nov '10

18 Nov '10

> Hi Patrick, > Sifting appender could probably help. Here is a sample config file: public class MyTest { Logger logger = LoggerFactory.getLogger(Foo.class); @Before public void setUp() { MDC.put("testname", "myTest"); } @Test a() { logger.info("a says hello"); ... } @Test b() { logger.info("b says hello"); ... } } public class MyOtherTest { Logger logger = LoggerFactory.getLogger(Foo.class); @Before public void setUp() { MDC.put("testname", "MyOtherTest"); } @Test x() { logger.info("a says hello"); ... } } >> Let us know if it works... >> -- Ceki Hi Ceki, It works except for the delete. When I call the test (class) MyTest (multiple times) logback don't delete the log file but append it. The first time when the Logger initialize the file is truncated but not for the second and more times until the Logger must initialize again. -Patrick

2 1