2 |
monit Bugs and To-Do Items |
monit Bugs and To-Do Items |
3 |
------------------------------------------------------------------ |
------------------------------------------------------------------ |
4 |
|
|
5 |
|
TODO: |
6 |
|
|
7 |
|
1.) add action for service monitoring disabling and enabling. New actions will |
8 |
|
support dependency and will be accessible through command line, such as: |
9 |
|
|
10 |
|
monit [monitor|unmonitor] httpd |
11 |
|
|
12 |
|
Tests supported actions in configuration file will be extended by unmonitor |
13 |
|
action, so it will be possible to write: |
14 |
|
|
15 |
|
if failed checksum then unmonitor |
16 |
|
|
17 |
|
|
18 |
|
2.) Timeout statement will be extended to provide generic framework for error |
19 |
|
ratio driven timeout. New syntax: |
20 |
|
|
21 |
|
if NUMBER EVENT within NUMBER cycles then timeout |
22 |
|
|
23 |
|
where: |
24 |
|
|
25 |
|
EVENT ... supported events (such as restart|timestamp|size|checksum|...) |
26 |
|
NUMBER ... integer |
27 |
|
|
28 |
|
It will be possible to let the service timeout (and don't flood the user |
29 |
|
with error messages) after specified limit is reached, such as: |
30 |
|
|
31 |
|
if 5 timestamp within 5 cycles then timeout |
32 |
|
if 3 restart within 5 cycles then timeout |
33 |
|
|
34 |
|
Following example is alternative to unmonitor action, with the exception |
35 |
|
that generic timeout is not inherited by dependants: |
36 |
|
|
37 |
|
if failed checksum then alert |
38 |
|
if 1 checksum within 1 cycle then timeout |
39 |
|
|