Now carries

• hpasm Hardware Checks
• hpacucli Raid Checks
• SNMP Connectivity, but does not use snmp

At the moment 'okconfig install' requires a password on the CLI. It would be good if it could prompt for password.

There are multiple issues with the new output from okconfig install.

1. Command crashes if running as user because user lacks write permissions to /var/log/okconfig/install_agent.log

2. When error occurs, no troubleshooting output is printed to the user, instead only the "please check install_agent.log for errors is outputted. This can get very annoying.

3. In some cases errors can occur and the only output in the error log is "please check install_agent.log for more information"

4. When installing from adagios and there are any errors, the only output is "permission denied" on install_agent.log

Steps to reproduce:

1. Install and configure OKConfig + adagios.
2. Using the adagios Web UI, create a 'mygroup' host group the OKConfig way.
3. Using the adagios Web UI, create a 'myhost' host the OKConfig way. Make it a member of the above group
4. Perform any action that involves Nagios verifying the configuration.

It produces the above error message. The host definition OKConfig created inherits from a 'linux-server' template o abstract host definition that isn't shipped either with OKConfig, pynag or adagios, nor with the Nagios default configuration.

define host {
    use                         linux-server
...

reading the OKConfig source code, this section shows other templates names 'windows-server' and 'generic-switch' are used that don't exist either.

I would like to be able to do something like this:
okconfig addtemplate <host> --template <t1>,<t2>,<t3>,<t4>
and have the four templates applied to the host.

This would save me the time writing for loops.

I'll write the patch; I just need to know where I should put the splitting logic (CLI script or in the library, CLI makes sense, right?). And also the preferred delimiter, it seems logical to use , but that would break support for templates with , in their name. I don't do that, have you seen it used somewhere? A possible workaround could be to check if the whole template string existed as a template and if not try to split it up and then finally fail if nothing succeeds.

Hostnames are missing at least

okconfig upgrade does not upgrade uses of sms notification commands.

Fix suggestion

pynag update service_notification_commands=okc-notify-service-by-sms where service_notification_commands=notify-service-by-sms
pynag update host_notification_commands=okc-notify-host-by-sms where host_notification_commands=notify-host-by-sms

okconfig install gives me a error referring /tmp/install_nrpe.sh which is default in py code.

/etc/okconfig.conf does not contain install_nrpe_script configuration variable on a older machine.

/etc/okconfig.conf.rpmnew was created during install because of local changes.

The opsview vmware plugins defined in the vmware template are not packaged but are defined in both services and commands configuration files.

A decision needs to be made on which plugins stay and which don't, what needs to be packaged and also some documentation on how to configure vmware monitoring.

If applicable okconfig upgrade should be run automatically when upgrading via rpm

I would like to see in a graph used capacity for storwize
maybe also total storage
default interval shouldn't be less than 15 minutes

nagios-okplugin-check_rhcs-0.0.4-1.el6.x86_64 config in /etc/nrpe.d/check_rhcs.cfg containt reference to
/usr/lib6464 instead of /usr/lib64

# cat /etc/nrpe.d/check_rhcs.cfg 
command[check_rhcs]=sudo /usr/lib6464/nagios/plugins/check_rhcs -c
command[check_rhcs_service]=sudo /usr/lib6464/nagios/plugins/check_rhcs -s '$ARG1$'

command[get_rhcs_fencing]=test ! -p /tmp/fence_manual.fifo
command[check_rhcs_fencing]=/usr/lib6464/nagios/plugins/check_rhcs_manualfencing.sh


command[get_rhcs_cman_group]=/sbin/group_tool ls '$ARG1$' '$ARG2$'
command[get_rhcs_cman_group_default]=/sbin/group_tool ls 0 default
command[get_rhcs_cman_group_rgmanager]=/sbin/group_tool ls 1 rgmanager


command[check_rhcs_cman_group]=/usr/lib6464/nagios/plugins/check_rhcs_cman_group.sh --level '$ARG1$' --group '$ARG2$'
command[check_rhcs_cman_group_default]=/usr/lib6464/nagios/plugins/check_rhcs_cman_group.sh --level 0 --group default
command[check_rhcs_cman_group_rgmanager]=/usr/lib6464/nagios/plugins/check_rhcs_cman_group.sh --level 1 --group rgmanager


# RHEL 6 specific
command[check_rhcs6_fencing]=/usr/lib6464/nagios/plugins/check_rhcs_fence

Warnings and critical are silently returned as OK into nagios

Following shows up green in Nagios

PROCS WARNING: 0 processes with command name 'crond'

--alias parameter does nothing

Creates a new nrpe.cfg but what happens if an nrpe.cfg is already there with something valuable?

usr/lib64 is hardcoded inside check_procs.sh

Traceback (most recent call last):
File "/usr/bin/okconfig", line 351, in
parse_arguments()
File "/usr/bin/okconfig", line 66, in parse_arguments
addcontact()
File "/usr/bin/okconfig", line 222, in addcontact
f = okconfig.addcontact(contact_name=opts.contact, alias=opts.alias, force=opts.force, group_name=opts.group, email=opts.email)
File "/usr/lib/python2.6/site-packages/okconfig/init.py", line 273, in addcontact
contact = pynag.Model.Contact.objects.get_by_shortname(contact_name)
File "/usr/lib/python2.6/site-packages/pynag/Model/init.py", line 382, in get_by_shortname
return ObjectFetcher._cached_shortnames[self.object_type][shortname]
KeyError: 'None'

Errors:
install_nrpe.sh: line 14: get_os_release: command not found

The OS on the target host is:

NAME="Ubuntu"
VERSION="12.04.1 LTS, Precise Pangolin"
ID=ubuntu
ID_LIKE=debian
PRETTY_NAME="Ubuntu precise (12.04.1 LTS)"
VERSION_ID="12.04"

But something tells me it's failing before that.

requirements.txt must contain paramiko and any other dependencies

The Windows installer agent hangs even after following the instructions in the wiki.

https://github.com/opinkerfi/okconfig/blob/master/docs/okconfig.pod

The URL http://opensource.is/trac/wiki/OKConfig returns a 404.

I used okconfig to set up a quick monitoring setup of 100 hosts. I did it like this:

for i in $(seq 1 100); do printf "okconfig addhost --host knc%04i --group knc-miners\n" $i; done > doit
for i in $(seq 1 100); do printf "okconfig addservice host_name=knc%04i use=okc-linux-check_ssh\n" $i; done > doit2

Sample from each file:
first file:

okconfig addhost --host knc0008 --group knc-miners
okconfig addhost --host knc0009 --group knc-miners
okconfig addhost --host knc0010 --group knc-miners
okconfig addhost --host knc0011 --group knc-miners
okconfig addhost --host knc0012 --group knc-miners

second:

okconfig addservice host_name=knc0008 use=okc-linux-check_ssh
okconfig addservice host_name=knc0009 use=okc-linux-check_ssh
okconfig addservice host_name=knc0010 use=okc-linux-check_ssh
okconfig addservice host_name=knc0011 use=okc-linux-check_ssh
okconfig addservice host_name=knc0012 use=okc-linux-check_ssh

then I just ran doit1 and then doit2 and I had what I needed to monitor these systems. I do however see an issue with the resulting config:

root@ds-0998:~# cd /etc/nagios3/okconfig/services/
root@ds-0998:/etc/nagios3/okconfig/services# ll
total 204
drwxr-xr-x 2 root root 4096 Nov 20 14:47 ./
drwxr-xr-x 5 root root 4096 Nov 6 18:33 ../
-rw-r--r-- 1 root root 197599 Nov 20 14:47 untitled.cfg

it contains:

Configuration file /etc/nagios3/okconfig/services/untitled.cfg

Edited by PyNag on Wed Nov 6 19:02:48 2013

define service {
use okc-linux-check_ssh
}

Configuration file /etc/nagios3/okconfig/services/untitled.cfg

Edited by PyNag on Wed Nov 6 19:02:48 2013

Values from templates:

check_command check_ssh

active_checks_enabled 1

obsess_over_service 1

is_volatile 0

process_perf_data 1

check_period 24x7

notification_interval 0

notification_period 24x7

failure_prediction_enabled 1

retain_status_information 1

event_handler_enabled 1

flap_detection_enabled 1

notification_options w,u,c,r

retry_check_interval 1

retain_nonstatus_information 1

notifications_enabled 1

contact_groups admins

service_description SSH Connectivity

parallelize_check 1

passive_checks_enabled 1

normal_check_interval 5

max_check_attempts 4

check_freshness 0

define service {
use okc-linux-check_ssh
host_name knc0001
}

This block is reapeated 99 more times, once for each host.

Environment:

Request Method: GET
Request URL: http://localhost:8000/okconfig/edit/tandoori

Django Version: 1.4.1
Python Version: 2.7.3
Installed Applications:
('django.contrib.auth',
'django.contrib.contenttypes',
'django.contrib.sessions',
'django.contrib.sites',
'django.contrib.messages',
'adagios.okconfig_',
'adagios.objectbrowser',
'adagios.rest',
'adagios.misc')
Installed Middleware:
('django.middleware.common.CommonMiddleware',
'django.contrib.sessions.middleware.SessionMiddleware',
'django.middleware.csrf.CsrfViewMiddleware',
'django.contrib.auth.middleware.AuthenticationMiddleware',
'django.contrib.messages.middleware.MessageMiddleware')

Traceback:
File "/usr/lib/python2.7/site-packages/django/core/handlers/base.py" in get_response

1.                     response = callback(request, _callback_args, *_callback_kwargs)
   

   File "/home/tommi/coding/adagios/adagios/okconfig_/views.py" in edit

2.     form = forms.EditTemplateForm(service=service,initial=initial)
   

   File "/home/tommi/coding/adagios/adagios/okconfig_/forms.py" in init

3.     self.command_line = service.get_effective_command_line()
   

   File "/usr/lib/python2.7/site-packages/pynag/Model/init.py" in get_effective_command_line

4.         command = Command.objects.get_by_shortname(command_name)
   

   File "/usr/lib/python2.7/site-packages/pynag/Model/init.py" in get_by_shortname

5.     return ObjectFetcher._cached_shortnames[self.object_type][shortname]
   

Exception Type: KeyError at /okconfig/edit/tandoori
Exception Value: 'check_nrpe'

Add /etc/sudoers.d/check_hpacucli with:

nrpe        ALL=(ALL)       NOPASSWD: /usr/sbin/hpacucli

Modify /usr/lib64/nagios/plugins/check_hpacucli.py
-sudo=False
+sudo=True

For security reasons I don't have remote root account enabled. It would be good if you could pass --sudo to the command line so that you can use a sudoer to remote in instead of root.

Not sure if you would need to have a line in sudoers for nopasswd or whether you can prompt for it via the new --interactive command?

Is it possible to tag a new version of okconfig? I've hit some issues that are always fixed in master :)

default certificate check looks like this:

/usr/lib64/nagios/plugins/check_http -H google.com --ssl -C 21,2

wrong repo, moving...

Right now a few of the okconfig templates have built in groups, it is inconsitent between templates whether they are there or not. We need to decide the "right way" (TM)

cisco/hostgroups.cfg:define hostgroup {
cisco/hostgroups.cfg-   name    cisco-hostgroup
--
hp/hpasm/groups.cfg:define servicegroup {
hp/hpasm/groups.cfg-    servicegroup_name   proliant-services
--
hp/hpasm/groups.cfg:define contactgroup {
hp/hpasm/groups.cfg-    contactgroup_name   proliant-contacts
--
hp/eva/groups.cfg:define servicegroup {
hp/eva/groups.cfg-  servicegroup_name   eva-services
--
hp/eva/groups.cfg:define contactgroup {
hp/eva/groups.cfg-  contactgroup_name   eva-contacts
--
misc/servicegroups.cfg:define servicegroup {
misc/servicegroups.cfg- servicegroup_name           windows-services
--
misc/servicegroups.cfg:define servicegroup {
misc/servicegroups.cfg- servicegroup_name   linux-services
--
mssql/servicegroups.cfg:define servicegroup {
mssql/servicegroups.cfg-    servicegroup_name   mssql-services

okc-check-generic-brocade-prod-health isn't defined anywhere.

Found something similar

###########################################################
# Check Hardware Health on Brocade SAN Switches
define service {
        name                    check-generic-brocade-health
        use                     generic-service

        service_description     Hardware Health Status
        check_command           check_brocade_san_env
        normal_check_interval   1
        retry_check_interval    1
        #servicegroups          brocade-san-switch-health
        _SNMP_COMMUNITY         public

        #action_url             /pnp4nagios/graph?host=$HOSTNAME$&srv=$SERVICEDESC$

        register                0
}

You can monitor the default instance (port 1433) but there is no way to add different instances

Add /etc/sudoers.d/check_hpasm

nrpe        ALL=(ALL)       NOPASSWD: /sbin/hpasmcli -s *

config.py doesn't parse the value for examples_directory_local:

a new line 47 should be enough to fix it:
elif keyword == "examples_directory_local": examples_directory_local = value

Hi,

Is the windows client install meant to work?

I just get this:

The system cannot find the file specified.
Error: Directory /usr/share/okconfig/client/windows//nsclient not found

Cheers

Should be contactgroups

To list available templates from the command line