[logback-dev] [JIRA] Created: (LBCORE-152) When your application is stopped and restarted, logging will not continue at the correct location, i.e. at the largest index number for the current period.

[David Gonzalez (JIRA)]

When your application is stopped and restarted, logging will not continue at the correct location, i.e. at the largest index number for the current period.
-----------------------------------------------------------------------------------------------------------------------------------------------------------

                 Key: LBCORE-152
                 URL: http://jira.qos.ch/browse/LBCORE-152
             Project: logback-core
          Issue Type: Sub-task
          Components: Rolling
    Affects Versions: 0.9.20
            Reporter: David Gonzalez
            Assignee: Logback dev list


As the online manual says (http://logback.qos.ch/manual/appenders.html) :
"When your application is stopped and restarted, logging will continue at the correct location, i.e. at the largest index number for the current period.".
This mean that the following steps occured :
- You have a log.trc current log file
- When rolling, as the <fileNamePattern> defined is <fileNamePattern>log_%i.log.zip</fileNamePattern>, archived log filed are created with the index %i starting at Zero.
- If current index is 3 when application is stopped, then restarting index is 4 and next archived log file to be created is log_4.log.zip

But this does not occure. Unfortunately the new archived log file with index equals to 4 is not created and you see a file with a weird name log_tmp<nanosseconds>.log created.
The reason has to be linked to LBCORE-131, current period index is not properly managed then when system is restarted this index is newly set to Zero and not 4.
As system tries to create log_0.log.zip file, system does test that this file exists and stops.
Temporary file tmp<nanos> which has to be saved as archived log file stays created (should have been deleted after archived log file creation successed).

As LBCORE-131 gives a fix, this fix should be improve in order to test also if current index period has to be increase because archived lof file in COMPRESS MODE already exists.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira