465 |
For variable value test types will monit send alert each time the |
For variable value test types will monit send alert each time the |
466 |
monitored object changed. This involves: |
monitored object changed. This involves: |
467 |
|
|
468 |
|
o Monit started, stopped or reloaded |
469 |
o A file checksum changed |
o A file checksum changed |
470 |
o A file size changed |
o A file size changed |
471 |
o A file content match |
o A file content match |
569 |
nonexist |
nonexist |
570 |
data |
data |
571 |
icmp |
icmp |
572 |
|
instance |
573 |
invalid |
invalid |
574 |
exec |
exec |
575 |
changed |
changed |
598 |
check process mybar with pidfile /var/run/mybar.pid |
check process mybar with pidfile /var/run/mybar.pid |
599 |
alert foo@bar only on { timeout } |
alert foo@bar only on { timeout } |
600 |
|
|
601 |
|
The 'instance' alert type report events related to monit internals, |
602 |
|
such as when monit instance was started/stopped/reloaded. |
603 |
|
|
604 |
|
When the alert mediator is unavailable (for example when mailserver |
605 |
|
failed), monit will queue the events until mediator recovery and |
606 |
|
post them in order with original timestamp so the events are not |
607 |
|
lost. |
608 |
|
|
609 |
|
|
610 |
=head2 Alert message layout |
=head2 Alert message layout |
611 |
|
|
668 |
are fixed and are: |
are fixed and are: |
669 |
Event: Failure state: Recovery state: |
Event: Failure state: Recovery state: |
670 |
|
|
671 |
EVENT_CHANGED "Changed" "Changed back" |
EVENT_CHANGED "Changed" "Changed back" |
672 |
EVENT_CHECKSUM "Checksum failed" "Checksum passed" |
EVENT_CHECKSUM "Checksum failed" "Checksum passed" |
673 |
EVENT_CONNECTION "Connection failed" "Connection passed" |
EVENT_CONNECTION "Connection failed" "Connection passed" |
674 |
EVENT_DATA "Data access error" "Data access succeeded" |
EVENT_DATA "Data access error" "Data access succeeded" |
675 |
EVENT_EXEC "Execution failed" "Execution succeeded" |
EVENT_EXEC "Execution failed" "Execution succeeded" |
676 |
EVENT_GID "GID failed" "GID passed" |
EVENT_GID "GID failed" "GID passed" |
677 |
EVENT_ICMP "ICMP failed" "ICMP passed" |
EVENT_ICMP "ICMP failed" "ICMP passed" |
678 |
EVENT_INVALID "Invalid type" "Type passed" |
EVENT_INSTANCE "Monit instance changed" "Monit instance changed not" |
679 |
EVENT_MATCH "Regex match" "No regex match" |
EVENT_INVALID "Invalid type" "Type passed" |
680 |
EVENT_NONEXIST "Does not exist" "Exists" |
EVENT_MATCH "Regex match" "No regex match" |
681 |
EVENT_PERMISSION "Permission failed" "Permission passed" |
EVENT_NONEXIST "Does not exist" "Exists" |
682 |
EVENT_RESOURCE "Resource limit matched" "Resource limit passed" |
EVENT_PERMISSION "Permission failed" "Permission passed" |
683 |
EVENT_SIZE "Size failed" "Size passed" |
EVENT_RESOURCE "Resource limit matched" "Resource limit passed" |
684 |
EVENT_TIMEOUT "Timeout" "Timeout recovery" |
EVENT_SIZE "Size failed" "Size passed" |
685 |
EVENT_TIMESTAMP "Timestamp failed" "Timestamp passed" |
EVENT_TIMEOUT "Timeout" "Timeout recovery" |
686 |
EVENT_UID "UID failed" "UID passed" |
EVENT_TIMESTAMP "Timestamp failed" "Timestamp passed" |
687 |
|
EVENT_UID "UID failed" "UID passed" |
688 |
|
|
689 |
=item * |
=item * |
690 |
|
|
793 |
the SMTP server. |
the SMTP server. |
794 |
|
|
795 |
|
|
796 |
|
=head2 Event queue |
797 |
|
|
798 |
|
Monit allows to queue the events which failed to deliver due to |
799 |
|
the alert handler error. For example in the case that no mailserver |
800 |
|
is available, monit will redirect the events to the queue and |
801 |
|
try to reprocess them next cycle. As soon as the alert handler |
802 |
|
(i.e. mailserver) recovered, monit will post the queued events. |
803 |
|
The queue is persistent across monit restarts and provided that |
804 |
|
the backend filesystem is persistent too, across system restart |
805 |
|
as well. |
806 |
|
|
807 |
|
By default is the queue disabled and when the alert handler fails, |
808 |
|
monit will drop the alert message. To enable the event queue, it |
809 |
|
is needed to add the following statement to the monit control file: |
810 |
|
|
811 |
|
SET EVENTQUEUE BASEDIR <path> [SLOTS <number>] |
812 |
|
|
813 |
|
The <path> is the path to the directory, where events will be |
814 |
|
stored. Optionaly if you want to limit the queue size (maximum |
815 |
|
events count), you can use the slots option. When the slots |
816 |
|
option is not used, monit will store as many events as the |
817 |
|
backend filesystem allows. |
818 |
|
|
819 |
|
Example: |
820 |
|
|
821 |
|
set eventqueue |
822 |
|
basedir /var/monit |
823 |
|
slots 5000 |
824 |
|
|
825 |
|
The events are stored in binary format, one file per event. |
826 |
|
The queued event file size is ca. 130B or more (mainly depending |
827 |
|
on the message length). The file name is composed of the unix |
828 |
|
timestamp, underscore and the service name, for example: |
829 |
|
|
830 |
|
/var/monit/1131269471_apache |
831 |
|
|
832 |
|
In the case that you are running more then one monit instance |
833 |
|
on the same machine, you should use separated queue directories. |
834 |
|
|
835 |
|
If you want to purge the queue by hand (remove queued events/files), |
836 |
|
monit should be stopped before the removal. |
837 |
|
|
838 |
|
|
839 |
|
|
840 |
=head1 SERVICE TIMEOUT |
=head1 SERVICE TIMEOUT |
841 |
|
|
842 |
B<monit> provides a service timeout mechanism for situations |
B<monit> provides a service timeout mechanism for situations |
2945 |
I<inode>, I<pid>, I<ppid>, I<perm(ission)>, I<process>, I<file>, |
I<inode>, I<pid>, I<ppid>, I<perm(ission)>, I<process>, I<file>, |
2946 |
I<directory>, I<device>, I<size>, I<unmonitor>, I<rdate>, |
I<directory>, I<device>, I<size>, I<unmonitor>, I<rdate>, |
2947 |
I<rsync>, I<data>, I<invalid>, I<exec>, I<nonexist>, I<policy>, |
I<rsync>, I<data>, I<invalid>, I<exec>, I<nonexist>, I<policy>, |
2948 |
I<reminder> and I<failed> |
I<reminder>, I<instance>, I<eventqueue>, I<basedir>, I<slot(s)> |
2949 |
|
and I<failed> |
2950 |
|
|
2951 |
And here is a complete list of B<noise keywords> ignored by |
And here is a complete list of B<noise keywords> ignored by |
2952 |
monit: |
monit: |