Manage Plugin 0.6.2 for monitoring networks, servers...

[caoboth]

I must admit on my manage_check.php its reporting the correct oid for uptime (this is for Mib not Mib2) im also running 0.6 and have some very strange times within manage.

Im going to keep checking ive just switched the manage system to use the snmp times rather than do an uptime poll internally which i have a look around.

do we also know if manage is continuing to be developed as there a load of extra addons/additions that could be made (e.g alerting windows / group specific alerts from up/down/reboot etc)

Andy

[streaker69]

I must admit on my manage_check.php its reporting the correct oid for uptime (this is for Mib not Mib2) im also running 0.6 and have some very strange times within manage.

Im going to keep checking ive just switched the manage system to use the snmp times rather than do an uptime poll internally which i have a look around.

do we also know if manage is continuing to be developed as there a load of extra addons/additions that could be made (e.g alerting windows / group specific alerts from up/down/reboot etc)

Andy

Yes, Gilles is still doing work with Manage, but I believe he has a real job somewhere so he works on it when he can.

If you look at the lines that I posted, the actual issue is with the MIB section, not the MIB2 section. MIB2 is actually correct, MIB is not, it's not pulling sysuptime, it's pulling hrsystemtime. I'm not sure why Gilles put the different values in for MIB and MIB2 as if you're looking for real system uptime, I believe you'd always want to pull sysuptime.

[danwdoo]

Hello,
I'm a little confused as to which script I should be using to monitor Windows services for uptime with Manage. I am running Cacti on Windows and many of the solutions in this thread list the scripts for Linux. Any clarification would help! I tried the updated Linux ones and they just return 0's so I'm guessing they aren't the right ones!

Thanks!
Danny

[mcutting]

Danny - search the Scripts and Templates forum for Windows Services (Cigamit) and Windows Processes (BSOD2600). Once you have these running, Manage detects these, and starts to monitor them.

[streaker69]

Danny - search the Scripts and Templates forum for Windows Services (Cigamit) and Windows Processes (BSOD2600). Once you have these running, Manage detects these, and starts to monitor them.

When you're adding a service to be monitored via SNMP, once you create the graph, do you need to re-index the services? That's what I've been having to do to make them work.

[steffenn]

1st of all, Manage 0.6 is a great plugin, good job

After establishing this plugin for the default admin user without problems, we get in trouble with user customizing:

Suse 10.3
PA 2.1
Cacti 0.8.7b
Manage 0.6

Despite setting all realm and permission flags in the userprofile of a 2nd user like for the admin user, the new user keeps unauthorized, when klicking the Manage Tab in thge Cacti GUI. What do we wrong?

Can I ask one further question: Are their an exchange plattform for Manage Plugin extensions, like templates and further probing of several device types.

Steffen

[laraujo07]

Hello,

I'm running cacti 0.87b in Debian Etch with PA 2.1, manage 0.6 and spine. The graphs was create well and all others plugins (thold, discovery...) too. The graphs about TCP, Windows Process and windows services are creating with success, not problem, but I was with problem in manage windows Services and windows process in the cases following.
1- When I was using the Cacti_poller and SNMP (performance) the windows process are without problems and all windows services are down.
2 - When I was using the Cacti_poller and Poller_output (performance) the windows process are up (with many false/positive and many email alerts UP/DOWN) and the windows services are up without problems.

Any ideas ? Thanks.

[streaker69]

If anyone is interested, I'm currently working on a modification to Manage that will do Text to Speech conversion and read the actual errors outloud using Flite (Festival Lite). So far, the only page I've needed to change to make this happen is manage_ajax.php. I should have something ready for others to test in a couple of days.

[apperrault]

That sounds like a "Bloom"ing great idea.

Wonders to self if the hint was too subtle...

It actually sounds like a great idea. Especially if you have a system that is dedicate to Cacti and it is always set at the Manage page. Kind of reminds me of the Monitor plugin that would use a voice prompt to alert you to a problem.

app

[streaker69]

That sounds like a "Bloom"ing great idea.

Wonders to self if the hint was too subtle...

It actually sounds like a great idea. Especially if you have a system that is dedicate to Cacti and it is always set at the Manage page. Kind of reminds me of the Monitor plugin that would use a voice prompt to alert you to a problem.

app

Bloom is almost ready, I think N3ncy is finishing up testing his script soon. We've had to make a couple of changes while we were testing things out.

And you hit exactly why I did this. I have my Cacti box sitting in my office and I have it logged in while I'm here with the Manage page up, and every 5 minutes it reads if there are any errors based upon the refresh timer built into Manage.

Of course, it will also read it on a remote workstation as I have Flite piping the output to a WAV file. On the local server, it reads it directly from the text file.

The only issue that I don't think I'm gonna be able to overcome is having it play the WAV file running on a Linux workstation running Firefox, as I haven't found a good solution for playing a WAV file embedded in a webbrowser on Linux. I'm open to suggestions if anyone has one regarding that.

Evil thought: Combine Flite with CactiCam. Sit at home and watch the office on CactiCam while SSH'd into the box. When you see someone enter the office type messages to be played over the cacti speakers addressed directly to them.

Of course, the Motion software that I'm using for Cacti can trigger events so it could be relatively automatic in speaking phrases, but it would be great if it could do face recognition and address people by name.

[pepj]

Hi Gilles
What about my changes ... agreeing or not ...

history:
1/It is very not funny when all selected parameters disappear (for example in "Event Reporting" the "filter" or "selected device") after a refresh or executing a request.
2/ It would be good at giving the number of the showed lines.

Here my corrections for this part:
in "manage_viewalerts.php" line 100:

Code: Select all

load_current_session_value("host_managed", "sess_host_managed", "1"); 
   load_current_session_value("event", "sess_event", "1"); 
   load_current_session_value("id", "sess_id", ""); 
   load_current_session_value("managenbrline", "sess_managenbrline", "20"); 

and for the number of lines of "Event Reporting"
"manage_viewalerts.php": near line 284

Code: Select all

               <td width="5"></td> 
<?php 
/* BEGIN CHANGES 2 */ 
      $managenbrline = (isset($_REQUEST["managenbrline"])) ? intval($_REQUEST["managenbrline"]) : 20; 
      $managenbrline = ($managenbrline<20) ? 20: $managenbrline; 
?> 
               <td width="20">&nbsp;&nbsp;&nbsp;&nbsp;nbr&nbsp;of&nbsp;events:&nbsp; 
            </td> 
            <td><input type='text' name='managenbrline' size='2' value='<?php print $managenbrline;?>'></td> 
<?php /* END CHANGES 2 */ ?> 

and "manage_viewalerts.php":

Code: Select all

$alerts = db_fetch_assoc("select 
      ida, 
      idh, 
      datetime, 
      ids, 
      message, 
      note, 
      oid 
      from manage_alerts 
      where idh=" . $_REQUEST["id"]. $tmp . " order by datetime DESC limit ".$managenbrline."");

[streaker69]

Ok, I think I have my modifications for the speaking portion of Manage ready for testing. I'll send it to the first three people that PM me here.

Pre-requisite though, you must have Flite already installed on your machine before you ask for it. Make sure that it works by creating a test wav file.

When I hear back from those three people that I send it to that it's working I'll publish it for everyone.

[gilles]

for the uptime problem, i include a patch from uptime plugin thread.
somebody was having a problem with a bad snmp agent.
you can adjust this in options :

Code: Select all

Check system uptime
If available, rather than snmp agent uptime.

[gilles]

Yes, which makes it inaccurate, which I have noticed before.

FOUND IT!

Read the info on the first image, then make the change from the second image. The wrong OID is being called resulting in the incorrect values for Windows boxes.

i think NO don't touch that.
just look down in the code you post, there is the correct oid (MIB II).

if you launch manage from command line and it's outputting it use MIB, i think you have a problem.
to use MIB II, which is more accurate, use options (see my precedent post).

if you could confirm that...

[gilles]

1/It is very not funny when all selected parameters disappear (for example in "Event Reporting" the "filter" or "selected device") after a refresh or executing a request.

i can't reproduce this.
can you explain me ?

currently, it's right that if you click save button, you come back to a clear screen.