The Community Forums

Interact with an entire community of cPanel & WHM users!
  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

11.26 Piped Logging Broken

Discussion in 'General Discussion' started by VeZoZ, Sep 29, 2010.

  1. VeZoZ

    VeZoZ Well-Known Member

    Joined:
    Dec 14, 2002
    Messages:
    248
    Likes Received:
    0
    Trophy Points:
    16
    cPanel Access Level:
    DataCenter Provider
    It appears as though 11.26 has broken the piped logging which was functional in 11.25.

    I've tested this with Apache and Litespeed both produce 0 byte domain.com-bytes_log files. So this appears to be a problem with the /usr/local/cpanel/bin/splitlogs and not either web server based on the fact both produce the same result.

    Obviously switching back to non piped logging fixes it but then we get many more web server restarts as a result.

    /usr/local/cpanel/bin/splitlogs is a compiled binary or I'd just fix it myself. So I'd hope you guys would consider fixing this and release the fix to all branches rather than waiting many months for it to be fixed in 11.28.
     
  2. GWade

    GWade Registered

    Joined:
    Oct 2, 2008
    Messages:
    2
    Likes Received:
    0
    Trophy Points:
    1
    11.26 Piped logging questions

    Thank you for bringing this to our attention.

    I haven't seen this behavior and it has not been reported in-house, so we need some further information in order to track down and solve the problem.

    1. Do you know when the problem first occurred? (version number or date of update)
    2. What OS is running where you see this symptom?
    3. Are you seeing it on multiple systems or just one?

    Probably the best way of dealing with the issue is to open a support ticket on the affected system. Our support team is very effective at helping uncover the information needed for development to track down and solve the problem.

    Thank you for your help and patience.
     
  3. VeZoZ

    VeZoZ Well-Known Member

    Joined:
    Dec 14, 2002
    Messages:
    248
    Likes Received:
    0
    Trophy Points:
    16
    cPanel Access Level:
    DataCenter Provider
    CentOS 5.5 was just after the update.

    Seems to be happening certain domains on servers. It's causing the entire bandwidth system to choke though best I can tell. You have errors like this:

    Code:
    [2010-10-01 08:58:00 -0500] warn [cpanellogd] Unable to delete '/usr/local/apache/domlogs/domain.com.bkup': No such file or directory
     at /usr/local/cpanel/Cpanel/Logs.pm line 764
            Cpanel::Logs::post_process_logs('delete', ARRAY(0xfcff560), HASH(0x109463c0)) called at /usr/local/cpanel/Cpanel/Logd.pm line 598
            Cpanel::Logd::dologs(HASH(0x109a7bb0), HASH(0x10a31ce0), '') called at /usr/local/cpanel/Cpanel/Logd.pm line 261
            Cpanel::Logd::scanlogs('') called at /usr/local/cpanel/libexec/cpanellogd line 109
    
    It looks as though the bandwidth processing breaks when encountering a non existent file. So it results in accounts having this happen to them:


    Code:
       224 -rw-r--r--   1 root   nobody      225188 Oct  1 13:27 domain2.com-bytes_log
       104 -rw-r--r--   1 root   nobody      100852 Oct  1 05:00 domain2.com-bytes_log.bkup
       116 -rw-r--r--   1 root   nobody      112625 Oct  1 01:15 domain2.com-bytes_log.bkup2
    

    I think I was therefor looking at this a bit wrong. The bandwidth log processing does not seem to handle missing files very well. So that explains why usage for all does not work.

    Running /scripts/runweblogs works on users where all bytes log files exist and have data. The ones that do not well this is what you get:

    Code:
    Log checker loaded ok..
    What user's logs did you want to update? username
    ==> cPanel Log Daemon version 24.0
    ==> Loaded RRDs: version 1.3006
    Processing user...
    Run Logs domain: username.com.com BW Limit: 31457280000 Domains: [aaa.username.com.com aaa.net bbb.username.com.com]
    info [safeapacherestart] Restart elapsed seconds: 2
    [2010-10-01 11:38:27 -0700] Process bandwidth for username
    Cannot load an empty file at /usr/lib64/perl5/site_perl/5.8.8/x86_64-linux-thread-multi/YAML/Syck.pm line 123.
    
    The user had all the log files existing. The issue was some of them were 0 bytes. As soon as it encounters one it seems to not work.

    Although very strange I managed to fill data on each log file.

    So I think the issue is not in the piped logging at all.

    Strange thing though once I got stats to every subdomain on the account it processed fine. After which it would process fine with 0 byte files going forward.

    I'm just not sure at this point because here's a single user account:

    Code:
        4 drwxr-x---   2 root   user      4096 Oct  1 09:54 user/
        4 -rw-r-----   1 nobody user       216 Oct  1 11:15 userdomain.com
        4 -rw-r--r--   1 nobody user        34 Oct  1 11:15 userdomain.com-bytes_log
    
    Code:
    ==> cPanel Log Daemon version 24.0
    ==> Loaded RRDs: version 1.3006
    Processing user...
    Run Logs domain: userdomain.com BW Limit: 62914560000 Domains: []
    info [safeapacherestart] Restart elapsed seconds: 2
    [2010-10-01 11:57:44 -0700] Process bandwidth for user
    Cannot load an empty file at /usr/lib64/perl5/site_perl/5.8.8/x86_64-linux-thread-multi/YAML/Syck.pm line 123.
    
    This was a newly created account. Run runweblogs again and it runs through this time.

    I'm confused might have to make a ticket on this one.
     
  4. VeZoZ

    VeZoZ Well-Known Member

    Joined:
    Dec 14, 2002
    Messages:
    248
    Likes Received:
    0
    Trophy Points:
    16
    cPanel Access Level:
    DataCenter Provider
    #4 VeZoZ, Oct 1, 2010
    Last edited: Oct 1, 2010
Loading...

Share This Page