forked from bucardo/check_postgres
-
Notifications
You must be signed in to change notification settings - Fork 0
/
TODO
74 lines (40 loc) · 2.41 KB
/
TODO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
Quick list of outstanding items / bugs / feature requests for CP:
NOTE: All bugzilla items are now on github
* Add --setup and --reset flags to replicate_row to automate things.
* Add documentation on privs required for each action, and recommended grants for roles, etc.
* The same_schema action does not check indexes. See bugzilla #54
* Perform automatic creation of views and function to allow all actions to be run
as non-superusers.
* Have the configuration file checker only report once on a changed checksum,
then store the changed one somewhere for the next run
* Put on CPAN - worth doing as this in now in packages?
* Fix up the constraints checking for same_schema. It sometimes reports odd
constraints with the name repeated many times.
* Ensure that all stats output is 100% Nagios certified,
e.g. 'dbname'=val1;warn;crit;extra
* Make sure all the regexes (e.g. slurp =~ //) work on WIN32.
* Add a --output=plain, for simple human-readable output.
* Add a --output=csv, for importing into various targets.
* Language testing: force to EN?
* Allow a flag to determine if skipped into goes into perf or not
* Make a (semi) automated way to allow non-superusers to run all actions
* Put complete Nagios templates on a page somewhere
* Get more language translations
* Investigate report of an empty fsm_pages causing 255 out of range error
* Check the current logfile is not bigger than X KB/MB/GB
* Have the checksum action only warn once and save the result somewhere
* Reconsider PGBINDIR and see if there is something more standard
* Support for pg_bouncer checks
* Full and correct support for Nagios, e.g.
http://nagiosplug.sourceforge.net/developer-guidelines.html#AEN203
* Test for all the Nagios output of each action.
* For last_analyze and last_vacuum, way to skip null (never been hit)
* Check at least one table has been vacuumed/analyzed since X time
* Check every table has been vacuum|analyze at least once. (no NULL)
* Check at least one table has been vacuumed|analyzed since X hours/days/...
* Check the current logfile is not bigger than X KB/MB/GB
* Make the bloat estimation more accurate (lots of small bad assumptions in there)
* Consider a more accurate bloat check, using the expensive pg_stattuple function
* For the Nagios 'perf' output on last_analyze and last_vacuum (and other things with a time),
consider making a single global timing call.
* Create a Perl::Critic test suite