Permalink
Commits on Oct 16, 2018
  1. Merge pull request #266 from naemon/readd_license_file

    jacobbaungard committed Oct 16, 2018
    Re-add COPYING file with license
Commits on Oct 10, 2018
  1. Merge pull request #265 from jacobbaungard/bugfix/MON-11295-always-ex…

    jacobbaungard committed Oct 10, 2018
    …ecute-checks-within-interval
    
    Always schedule next_check within check_interval
  2. Re-add COPYING file with license

    jacobbaungard committed Oct 10, 2018
    This commit adds a COPYING file with the GPLv2 license. This ensures
    that the we do not get a wrong license when running automake, and also
    that GitHub automatically can detect the license for the project.
    
    Signed-off-by: Jacob Hansen jhansen@op5.com
  3. Always schedule next_check within check_interval

    jacobbaungard committed Oct 10, 2018
    After #259 we now keep the
    next_check schedule over restarts if use_retained_schedule_info is
    enabled. However after this patch, if one would lower the check_interval
    it was possible that after the restart, the next check of an object
    would be more than one check_interval away.
    
    This commit ensures that if the next_check is more than one
    check_interval away, then we randomly schedule the next check, instead
    of using the retention data.
    
    This fixed MON-11295 (https://jira.op5.com/browse/MON-11295)
    
    Signed-off-by: Jacob Hansen <jhansen@op5.com>
Commits on Oct 8, 2018
  1. Merge pull request #257 from jacobbaungard/bugfix/MON-10565-cleaner-s…

    jacobbaungard committed Oct 8, 2018
    …hutdown
    
    Init: Increase delay between SIGTERM and SIGKILL
Commits on Oct 5, 2018
  1. fix pending dependencies

    sni committed Oct 5, 2018
    right now we did only check the state during dependency checks. But for pending hosts
    and services the state is usually OK/UP so the check passed. For pending flag checks
    we have to look at the has_been_checked flag as well. This leads to the situation where
    services checks will be run if the master service is in pending state even if the service
    has pending service execution failure flag set.
Commits on Oct 2, 2018
  1. el6: use correct logrotate script

    sni committed Oct 1, 2018
    Since we install the el7 logrotate in our Makefile.am without further OS
    detection we need to replace the logrotate file for el6 later. Otherwise we
    would end up with the el7 file and no logrotation.
Commits on Oct 1, 2018
  1. Merge pull request #259 from jacobbaungard/bugfix/MON-10720_retain-ne…

    jacobbaungard committed Oct 1, 2018
    …xt-schedule
    
    Retain next_check schedule on restart (#224, #156)
Commits on Sep 11, 2018
  1. Tests: Retain next_check schedule on restart

    jacobbaungard committed Sep 11, 2018
    This commit adds tests to ensure that the next_check is set correctly
    after Naemon restarts. This ensures the logic is from the previous
    commit is correctly followed.
    
    This fixes:
    - #224
    - #156
    - MON-10720 (https://jira.op5.com/browse/MON-10720)
    
    Signed-off-by: Jacob Hansen <jhansen@op5.com>
  2. Retain next_check schedule on restart

    jacobbaungard committed Sep 10, 2018
    This commit ensures that the next_check schedule for hosts and services
    are retained on Naemon restart, given that use_retained_scheduling_info
    is enabled.
    
    The logic is as follows:
    
    - If use_retained_scheduling_info is disabled, set a random time (as
      before)
    - If use_retained_schedule_info is enabled:
      - If we didn't miss the check during the restart, retain the old
        next_check time
      - If we missed one check, schedule the service/host within the next
        interval_length (usually 60 seconds)
      - If we missed more than one check, schedule the next check randomly.
    
    We schedule missed checks within 60 seconds, rather than immediately in
    order to do some load balacing. This is also the rationale for
    scheduling the check randomly, in case we missed more than one check
    (this indicates Naemon has been down for a longer period of time).
    
    This fixes:
    - #224
    - #156
    - MON-10720 (https://jira.op5.com/browse/MON-10720)
    
    Signed-off-by: Jacob Hansen <jhansen@op5.com>
Commits on Sep 4, 2018
  1. Init: Increase time till SIGKILL is sent

    jacobbaungard committed Sep 3, 2018
    When using killproc to kill a process, first a SIGTERM signal is sent,
    and after a default of 3 seconds (5 seconds on suse), a SIGKILL is sent.
    
    On larger setups we often see that 3 seconds, is not sufficient for
    Naemon to shutdown, and as a result retention data might not be
    correctly saved.
    
    This commit increase the timeout to 90 seconds on rehat based releases
    and on SUSE. This micics the behaviour of systemd.
    
    This is part of MON-10565
    
    Signed-off-by: Jacob Hansen <jhansen@op5.com>
  2. Log successful save of retention data

    jacobbaungard committed Sep 3, 2018
    In able to aid debugging in case Naemon was uncleanly shutdown.
    
    This is part of MON-10565
    
    Signed-off-by: Jacob Hansen <jhansen@op5.com>
Commits on Aug 29, 2018
  1. Merge pull request #255 from jacobbaungard/bugfix/MON-11112-fix-doubl…

    jacobbaungard committed Aug 29, 2018
    …e-newline-escape
    
    parse_check: Don't escape already escaped newlines
  2. No on-demand host checks when service is in hard critical

    jacobbaungard authored and sni committed Aug 24, 2018
    The current Naemon documentation states that on-demand host checks are
    performed when a service associated with the host changes state.
    This is not actually what happens, currently on-demand host checks
    are scheduled on any service checks which results in a NON-OK state for
    the service.
    
    This commit changes that logic slightly, so that on-demand host checks
    are performed when the service changes state OR if the service is in a
    soft state. That is, we no longer do on-demand host checks on non-ok
    service checks, if the service is already in hard critical.
    
    This ensures that we do not do a ridiculous amount of host checks when a
    host and all services are in hard critical.
    
    This fixes MON-5625
    
    Signed-off-by: Jacob Hansen <jhansen@op5.com>
  3. restore the strangely absent backticks

    pvdputte authored and sni committed Aug 28, 2018
  4. Reload : Fix defunct

    pvdputte authored and sni committed Aug 28, 2018
    Fixes issue #150 for debian/ubuntu packaging as well.
Commits on Aug 28, 2018
  1. parse_check: Don't escape already escaped newlines

    jacobbaungard committed Aug 28, 2018
    This patch ensures that newlines which are already escaped in a check
    results long_output is not escaped again.
    
    This fixes MON-11112
    
    Signed-off-by: Jacob Hansen <jhansen@op5.com>
Commits on Aug 3, 2018
  1. Update control

    sni committed Aug 1, 2018
    naemonstats has been moved to naemon-core and naemon-tools does no longer exist. Add replace/conflict tags for smooth updates.
Commits on Jul 17, 2018
  1. Make naemon-core depend on libnaemon

    RikaDenia authored and sni committed Jul 17, 2018
    Resolves upgrade issues when updating from 1.0.6 to 1.0.8.
Commits on Jul 16, 2018
  1. release 1.0.8

    sni committed Jul 9, 2018
  2. Merge pull request #248 from sni/add_missing_debug_newlines

    sni committed Jul 16, 2018
    add missing newlines to log_debug_info (fixes #247)
Commits on Jul 13, 2018
  1. add missing newlines to log_debug_info (fixes #247)

    sni committed Jul 13, 2018
    newlines are not automatically added and must be manually appended to each log entry.
    
    Signed-off-by: Sven Nierlein <sven@nierlein.de>
Commits on Jun 5, 2018
  1. Merge pull request #242 from sni/add_obsoletes

    sni committed Jun 5, 2018
    add obsoletes for naemon-tools
  2. add obsoletes for naemon-tools

    sni committed Jun 5, 2018
    naemon-tools are now part of the naemon-core project. Adding an obsoletes should
    make updates smoother.
    
    Signed-off-by: Sven Nierlein <sven@nierlein.de>
  3. Merge pull request #241 from wAmpIre/patch-1

    sni committed Jun 5, 2018
    daemon-init.in: Source "/etc/default/naemon" if exists
  4. daemon-init.in: Source "/etc/default/naemon" if exists

    wAmpIre committed Jun 5, 2018
    Set variables for naemon process (e.g. LANG and LC_ALL) to work around some (plugin threshold) issues
  5. Merge pull request #240 from naemon/fix_daemon_init_reload

    sni committed Jun 5, 2018
    Fix reload command on el6 (fixes #239)
  6. Fix reload command on el6

    sni committed Jun 5, 2018
    The backticks are missing, otherwise it trys to execute the pidfile.
Commits on Jun 1, 2018
  1. Merge pull request #237 from sni/bump_version

    sni committed Jun 1, 2018
    bump version in spec and debian files
  2. Merge pull request #236 from sni/release_1_0_7

    sni committed Jun 1, 2018
    Release 1 0 7
  3. release 1.0.7

    sni committed Jun 1, 2018
    Signed-off-by: Sven Nierlein <sven@nierlein.de>
  4. Merge pull request #234 from sni/fix_packaging

    sni committed Jun 1, 2018
    Fix packaging
  5. rpm: fix sles packaging

    sni committed Jun 1, 2018
    - sles12 does not allow files in /etc/sysconfig
    - sles11 needs to run autoreconf twice when there is no configure yet