New log rotation problem

I have resolved some log rotation problem...I have set del log rtation by size...and now the first time the log rotate,but then the output continue to go on the lod with the extension .log001 and not on the configured file.
why?

no,it the server rotates the log only the first time but then it continues to write on the .log001.On the original file it continues to write some server exception.
the server launch line in the scritp is the follow
nohup "$JAVA_HOME/bin/java" ${JAVA_VM} ${JAVA_OPTIONS_NODE} ${JAVA_OPTIONS} \
Dweblogic.Name=${SERVER_NAME} \
Dweblogic.ProductionModeEnabled=${PRODUCTION_MODE} \
Djava.security.policy="${WL_HOME}/server/lib/weblogic.policy" \
Dweblogic.Stdout=/u02/bea/weblogic81/mydomain/mydomain_LOG/myserver_Admin.log \
Dweblogic.Stderr=/u02/bea/weblogic81/mydomain/mydomain_LOG/myserver_Admin.log \
weblogic.Server > /dev/null &
and the config.xml part for the server configuration is the follow
<Server COMEnabled="true" CompleteMessageTimeout="0"
ConsoleInputEnabled="false" EnabledForDomainLog="true"
ExpectedToRun="false" JDBCLogFileName="./mydomain_LOG/jdbc.log"
JDBCLoggingEnabled="true" JavaCompiler="javac"
ListenAddress="10.2.1.16" ListenPort="10001"
MaxMessageSize="100000000" Name="myserver_Admin"
ServerVersion="8.1.3.0" SocketReaderTimeoutMaxMillis="10"
StagingMode="nostage" StdoutDebugEnabled="true"
StdoutEnabled="true" StdoutSeverityLevel="16" TransactionLogFilePrefix="../myserver_LOG/myserverAdminTransaction.log">
<SSL Enabled="false" IdentityAndTrustLocations="KeyStores"
ListenPort="10002" Name="myserver_Admin"/>
<Log FileCount="4" FileMinSize="30"
FileName="/u02/bea/weblogic81/mydomain/mydomain_LOG/myserver_Admin.log"
FileTimeSpan="1" Name="myserver_Admin"
NumberOfFilesLimited="true" RotationType="bySize"/>
<WebServer LogFileCount="2" LogFileLimitEnabled="true"
LogFileName="./mydomain_LOG/access.log"
LogRotationPeriodMins="360" LogRotationType="size"
MaxLogFileSizeKBytes="20000" Name="myserver_Admin"/>

Similar Messages

  • LMS3.2 Log rotation problem

    Hi,
    we use LMS3.2 (windows single-server)  for ~4500 devices and have a problem with log rotation.
    That's our configured log roation:
    Name with location
    Size (kb)
    Format
    No.of Backups
    1.
    D:\CSCOpx\log\dcrclient.log
    102400
    gz
    3
    2.
    D:\CSCOpx\log\ICServer.log
    102400
    gz
    3
    3.
    D:\CSCOpx\log\syslog.log
    1024000
    gz
    5
    The backup directory is the same as the usual log directory. The size of the 3 files is already bigger than the configured size in log roation, see:
    dcrclient.log     2,1 GB
    ICServer.log     1,1 GB
    syslog.log        9,7 GB
    After the log roation configuration we restarted the daemon manager, but it's not working. There are no backup files created and the 3 files are still growing.
    Regards
    Hendrik

    After you configure the log rotation you need and restart the daemons, you need to schedule the job. The schedule buttong is available under
    Common Services -> Server -> Admin -> Log Rotation.

  • Log rotation problem

    We are running WLS 8.1.4 and have had a recurring problem with log rotation for quite a long time. It has happened on every machine we've used, running both Windows 2000 and XP. What we want to do is have our logs roll nightly, but it seems that no matter what time we set in the WLS console, the log ends up rolled at an arbitrary time.
    Our current configuration is the following:
    Server File Name: ourServerName_%MM%-%dd%-%yyyy%.log
    Rotation Type: By Time
    Rotation Time: 04:00
    File Time Span: 24 Hrs
    We have tried many permutations on the Rotation Time field, including removing or adding leading zeros as well as seeing if the time difference is constant due to a time zone issue, but no discernible pattern has emerged.
    For instance, the logs for the last few days have been rolled at exactly 2300, despite being set to roll at 0400. Other times it has been set to roll at 0204, and it will roll at 1104. One other thing to note: the minutes are always correct, but the hours are nearly always off.
    Any ideas?

    No, and I can verify that this problem has showed up at multiple sites configured by multiple people besides myself. As we are precisely following the specs laid out in the WLS 8.1 docs, I can only assume this is a bug.

  • Cron-based log rotation problem

    In setting up Sun Web Server 6.1 2005Q4 SP5, I successfully restarted schedulerd after setting up cron-based log rotation for 12 AM for each Web instance. But still no rotated logs. I notice these messages in /app/sunone/https-admserv/logs/scheduler.error:
    Tue Mar 14 00:00:00 2006: Warning: cron has no information about finished child process 11303
    Tue Mar 14 23:00:03 2006: Warning: Process 11303 didn't finish in time, had to terminate it
    Any ideas?
    Thanks!

    This bug is addressed in SunSolve:
    Document ID:     4953147
    Title:     cron based log rotation fails when admin user is root and instance is non-root
    The workaround is to update this line in WS_ROOT/https-admserv/config/scheduler.conf:
        User <non-root user ID such as "nobody">to:
        User rootThen I restarted cron from the Server Manager to make sure the change was picked up.

  • Secure.log and log rotation problems

    I had originally posted this in the Tiger section, but now finding the Unix section which I think would be able to help me more. here is the link to the post: http://discussions.apple.com/thread.jspa?threadID=1913487&tstart=0
    Also I am getting this in my system.log, I have googled it and cannot find anything that tells me what it is:
    Feb 19 18:54:34 SamiMac kernel[0]: m378d97c0 0
    some help on this would also be appreciated.
    Message was edited by: a Mac user

    The secure.log message is explained at the bottom of Spurious Permission errors in OS X 10.4.
    For your ipfw logs, check that /etc/periodic/weekly/500.weekly contains the line
    for i in ftp.log lookupd.log lpr.log mail.log netinfo.log hwmond.log ipfw.log ppp.log secure.log; do
    (although I don't know why it wouldn't) and then check this is actually being run by looking at /var/log/weekly.out. Try
    tail -10 /var/log/weekly.out
    Let us know what you see.
    I haven't seen that kernel message before

  • [logging] Problems to configure logging rotation.

    Hi,
    I have an application .ear deployed in weblogic v10.3.1.0
    This application use java.util.logging to write a log file.
          fh = new FileHandler(logFileName,0,1,true);
          fh.setFormatter(new XMLFormatter());
          logger.addHandler(fh);
    FileHandler(String pattern, int limit, int count, boolean append)
    pattern - the pattern for naming the output file
    limit - the maximum number of bytes to write to any one file. If this is zero, then there is no limit. (Defaults to no limit).
    count - the number of files to use
    append - specifies append mode
    http://www.javadocexamples.com/java/util/logging/java.util.logging.FileHandler.html
    logFileName is dynamic with date formated like this yyyMMdd + ApplicationName + ".log"
    This file is created but I have also yyyyMMddSEC.log.1, yyyyMMddSEC.log.2, yyyyMMddSEC.log.3,...
    I DON'T WANT THESE FILES_, that's why I put limit to 0, count to 1 and append to true.
    This code works without jdev/weblogic but has not effect in weblogic.
    Q1. Why?
    So I go to Weblogic console: Domain Structure-> DefaultDomain->Logging
    Log file name: logs/DefaultDomain.log
    Rotation type: None
    NONE
    Messages accumulate in a single file.
    You must erase the contents of the file when the size is too large.
    Note that WebLogic Server sets a threshold size limit of 500 MB before it forces a hard rotation to prevent excessive log file growth.
    But it doesn't work, Weblogic continue to create log files like this *<filename>.log.<n>*
    Q2. Why?
    I have also created weblogic.xml in ViewControler/WEB-INF
    thanks to this documentation:
    http://download.oracle.com/docs/cd/E13222_01/wls/docs103/webapp/weblogic_xml.html#wp1063199
    but it doesn't work...again.
    Q3. Why?
    Q4. If I want applications manage themselves their log, how to deactivate the logging handler in weblogic (LogFileMBean?)
    Thanks for your help.

    You may want to ask in the WebLogic Server - Diagnostics / WLDF / SNMP forum. They own logging.

  • Problems happened when I changed my Apple ID to a new log in. Aperture is no longer on the MAC.

    Several months ago I change my Apple ID but not any of the hardware and products associated with it. I purchased Aperture under the old log in and have been using it regularly on my MAC. Last week my harddrive on the MAC crashed. When I went to reload all the purchases - everything under the new log in loaded back fine but everything from the old ID was no longer available. Aperture is gone...
    Attempts to log in using the old log in information has failed as well as countless hours working with customer service.
    I will never change my ID again - but that is not helping now.
    Does anyone have any other suggestions?

    Jim,
    are you sure, that you bought Aperture from the App Store, and not from the Apple Store?
    If you bought from the Apple Store, you will have received a serial number. Do you still have it? Then download the Aperture 3 Trial and reinstall - but do this only, if you have a serial number!
    Go to the Applications folder and delete Aperture.
    Download the Aperture 3.1 Trial  from here: Aperture 3.1 Trial
    Open the Aperture Trial disc image.
    Double-click the "Aperture Trial.mpkg" file.
    Follow the onscreen instructions to install Aperture.
    After the installation completes, choose Software Update from the Apple () menu.
    Install the Aperture 3.4 update from the list of available updates.
    If you definitely bought from the AppStore and not AppleStore - what happens when you try to log into the AppStore - in what way does it fail?
    Regards
    Leónie

  • Automatically log me out when create a new Ad Rotator

    Please help. I am not able to create a new Ad Rotator. When I am trying to create one, it shows the message below in 1 seconds, and direct me to the login page. I am so frustrated.

    Hi Gaurav:
    Thank you so much.  I disabled the adblock and  it is working now

  • Regarding internal daemon log rotation...

    I found a problem with internal daemon log rotation in iplanet webserver.it is not happening properly..........how to troble shoot this issue?

    Web Server 6.0 is EOSLed. Please move to the newer Web Servers 6.1 or 7.0. Check [http://www.sun.com/download/index.jsp?cat=Web%20%26%20Proxy%20Servers&tab=3&subcat=Web%20Servers|http://www.sun.com/download/index.jsp?cat=Web%20%26%20Proxy%20Servers&tab=3&subcat=Web%20Servers]. If you still have problems in these(6.1 or 7.0) versions of Web Server let us know.

  • Log Rotation on iPlanet 6.0 SP6

    Hello,
    I am running iPlanet 6.0 SP 6 on a Windows 2000 server and am having an issue with the access logs rotating. I have them set to rotate at 12:00 am using the Internal Daemon. The logs will rotate and begin to write to a new access file, however, there is no data in the log file. All that is showing up is the first line and the size of the log file is 1 KB. If anyone else has had this problem, please let me know. Thanks

    I believe this may happen when a program, such as ntpdate, changes this system clock by more than a few seconds. The problem has been seen when the system clock is particularly slow and the clock is synchronized by a nightly cron job. Under such circumstances, the access log file may be rotated twice, the second rotation overwriting the first. Bug 6206179 has been filed to track this problem.
    A work around is to modify the flex-rotate-init line in your magnus.conf configuration file so that it reads as follows:Init fn="flex-rotate-init"
         rotate-start="..."
         rotate-interval="..."
         rotate-suffix=".%Y%m%d%H%M.tmp"
         rotate-callback="/home/webservd/postrotate"Where /home/webservd/postrotate is an executable script that contains the following:#!/usr/bin/ksh
    tmpfile=$1
    logfile=${tmpfile%.tmp}          # removes .tmp part
    if [[ -s $logfile ]];            # if logfile exists and not empty
    then
        cat $tmpfile >>$logfile      # append the content of tmpfile to the logfile
        rm $tmpfile
    else
        mv $tmpfile $logfile
    fiThis will keep the server from overwriting access log entries if it tries to rotate the access log file twice.

  • Log Rotation

    Hi,
    We are monitoring our logs with several scripts, however we are finding inconsistency with the time of new logs created like access and error. We would like to have it created a new file everyday at 12:01 AM and should end at 11:59 PM.
    Please let me know if someone knows how to do it with some DS Settings.
    Thanks, John.

    These are the links to the official docs for 5.x:
    http://download.oracle.com/docs/cd/E19341-01/817-5221/logs.html
    From this manual you see that the only way to rotate by hand the log files is:
    1) Stop the Directory Server instance
    2) rename current access/error/audit log files
    3) Restart Directory Server instance
    In that case the DS will create a new access/error/audit log file.
    Moreover from:
    http://download.oracle.com/docs/cd/E19341-01/817-5235/config.html#wp21217
    We see that in 5.x version the attribute nsslapd-accesslog-logrotationtime has a different syntax (single integer value) compared to what it has in 6.x and later releases, that doesn't allow to specify you the exact time to rotate the log files. It will simply rotate after the amount of units expressed by the nsslapd-accesslog-logrotationtimeunit attribute.
    Given that we could mix a bit of both, and set a rotation policy from the Administration Console of the Directory Server Instance->Configuration Tab->Logs:
    - Check the box Unlimited size
    - Create new log every: *1 Day(s)*
    - Click Save
    A pop-up box will inform you that:
    You have set an unlimited size for the log files. Note that this can be dangerous as it can cause problems to your system if the size of the logs is very big. Do you want to continue?
    You can acknowledge this message and click Yes
    This should create a policy to rotate the access log file every 1 day ( = 23h59m59s after last log rotation) regardless of log size. The only advice I have with these settings is that you've to carefully monitor also the filesystem occupation, or set a proper deletion policy.
    Last step is then to sync the log time rotation, you have to 'manually' perform a rotation at midnight
    HTH,
    marco

  • GC log rotation question

    Hi,
    So after JDK 1.6.0_34 (as far as I know), java introduced GC log rotation JVM args (see below)
    -XX:+UseGCLogFileRotation -XX:NumberOfGCLogFiles=10 -XX:GCLogFileSize=1M
    Prior to these JVM arg additions, on any java application start, your GC log would get wiped. In my case a tomcat server running java code.
    So now I would assume when these JVM arg's are present, the GC log continues to rotate and not get wiped on application start up. However the above behavior still persists ever after JDK 1.6.0_34 with the new JVM args present.
    I should mention while the java application is running, the GC log DOES rotate (i.e., gc_log.txt.0, becomes gc_log.txt.1), but when application is restarted the LAST GC log is erased completely.
    This behavior isn't really "log rotation/backup", I've had to introduce my own shell/bat scripts to rotate the latest GC logs manually outside of these JVMargs. I'm not sure if I'm misunderstanding how to use the new JVM args or if oracle doesn't consider this a bug and this is normal behavior.
    I did file a bug on this, 9005051. However there is no guarantee your bug request ever makes it to the public bug database and oracle doesn't provide reasoning back to the submitter when the bug is rejected or not looked at, so I don't really know what happened with my bug request.
    If anyone has any insight into this, I would really appreciate it.
    Thank you!

    > Or better yet, where I can find someone to actually answer my question?
    Me! Me! I know the answer to that question...
    Oracle is more than happy for you to pay them for a support\service contract of which part of that will be that you can ask them questions and they will answer them.  Real Oracle engineers and perhaps the very people who created the code in the first place.  Or at least the ones in charge of it now.
    > yet there is no guarantee my bug is even a bug or will even get posted to the bug database
    Yes that is true.
    But I suspect, strongly, that if you have a big enough contract that every bug you post to the database, and that you tell oracle about, will appear promptly and be addressed promptly.
    And if you have a smaller (but not really small) Oracle contract I am rather certain that you can obtain a custom fix.
    Alternatively you can download the source, figure out the problem, and fix it yourself.  Then apply it to your running services.

  • Access and Error Log Rotation

    When turning on Archive for the logs. cron style. The log gets archived with the date etc, but the new files access and error do not get created. I have ns-cron on and the rotation occurs, but the files name only changes and it never create the new files and only keeps changing the file names with the new name.

    You may be running into Problem 4684892 mentioned in the release notes:
    http://docs.sun.com/source/817-5170-10/rn60sp7.html
    The Administration Server and the cron daemon must be run as root for cron-based log rotation to function properly. You may have to modify your cron.conf user to run as root.
    Thanks,
    Manish

  • Cisco Security manager syslog.log file problem

    Hello
    I have this problem with the CSM, the next file Syslog.log  (C:\Program Files\CSCOpx\log\Syslog.log  ), this file grows very fast to fill the hard disk and saturates the server, I have tried the log rotation of the cisco works but it doesnt work, what else can i do?
    the hard drive fills in 4 hours. tankyou

    In CSM clinet under Tools > CSM Administration > Debugging you can changing the level to something higher than debugging.
    I hope it helps.
    PK

  • Oracle 10G Checkpoint not complete Cannot allocate new log

    We have an oracle 10G database that has 4 groups of 200M redo logs. We are constantly seeing Checkpoint not complete followed by Cannot allocate new log.. messages in the alert log. When I look at the v$instance_recovery table I see this:
    SQL> select WRITES_MTTR, WRITES_LOGFILE_SIZE, WRITES_LOG_CHECKPOINT_SETTINGS, WRITES_OTHER_SETTINGS,
    2 WRITES_AUTOTUNE, WRITES_FULL_THREAD_CKPT from v$instance_recovery;
    WRITES_MTTR WRITES_LOGFILE_SIZE WRITES_LOG_CHECKPOINT_SETTINGS
    WRITES_OTHER_SETTINGS WRITES_AUTOTUNE WRITES_FULL_THREAD_CKPT
    0 0 66329842
    0 309461 41004
    It seems most of our checkpointing is being caused by the log_checkpoint_interval being set to 10000, which means we are doing 20 checkpoints during one redo log since our OS blocksize is 1024. What I'm thinking about doing is first adding two more redo log groups, and then either increasing the log_checkpoint_interval to 40000 (5 checkpoints for one redo log) or unsetting log_checkpoint_interval and setting fast_start_mttr_target to 600. Which would be the recommended parameter change? I think Oracle recommends using the fast_start_mttr_target. Are there any other suggestions?

    Hi
    >>unsetting log_checkpoint_interval and setting fast_start_mttr_target to 600.
    its a better idea,look on it
    http://download-uk.oracle.com/docs/cd/B19306_01/server.102/b14237/initparams068.htm#REFRN10058
    Just set fast_start_mttr_target parameter.it will resolve ur problem.
    Thanks
    Kuljeet

Maybe you are looking for