NetXMS Support Forum

Quote from: Tursiops on February 22, 2017, 02:45:53 AM
Not sure if anyone else has seen something like this?

I'm using to LOGWATCH subagent to monitor both standard windows event log and custom event logs (I have one created by my application) and up to know I didn't notice that kind of behavior.

I'm currently on version 2.0.8 on both server and agents.

Marco.

Quote from: Victor Kirhenshtein on January 18, 2017, 12:17:12 AM
Hi,

it cannot be solved without change in code. Please register issue in bug tracker to make Y axis always started at 0 configurable option for line charts.

Best regards,
Victor

Hi Victor,

registered as new feature request:

https://dev.raden.solutions/issues/1402

Thank you very much for your support.

Regards,
Marco

Quote from: Marco Incalcaterra on January 06, 2017, 08:19:04 AM

Quote from: Victor Kirhenshtein on January 05, 2017, 07:29:30 PM
Hi,

there is no way to determine exact cause without access to agent log. In agent log you will find record started with text "LogWatch: Unable to open event log". Btw, if you can edit agent's config from console, you can enable file manager subagent and download agent's log file.

Best regards,
Victor

Unfortunately no access even for the config file (agent is configured to reply only to "Servers"), I'll ask for a copy of the log file to analyze it. Thank you.
Regards,

Marco

Problem solved, just for future reference: it is a virtual machine with automatic deployment, the specific Event Log monitored (custom) was not yet created when the agent started. Just changed the post deployment step to create the custom Event Log before starting the agent.

Regards,
Marco

Quote from: Tursiops on January 17, 2017, 12:04:34 AM
Hi Marco,

You can right click on the graph, select Properties, then change the Y Axis Range.
However, I believe this is limited to 2^31 as the highest value it will accept, if you enter something higher, the results will not be as expected. Not sure if that's only if you're using 32bit Java and Console. Haven't checked with the web interface.

Alternatively you can simply zoom in on the part you want to see in detail (I think that only works in the console, not the web interface).

Hope that helps

Hello Tursiops,

it is a dashboard with data counting continuously (always increasing in this case), I cannot set the Y axis to a specific range. With automatic set I was expecting a kind of window "adjusted" to the vertical extent of the current data fitted into the window.

Regards,
Marco

Hi!

is the Y axis always starting from 0 (with positive values)? In auto scale mode seems to start always from zero, if you have small variation in the data plotted, it is difficult to evaluate the differences, line seems almost flat.

Is there a way to avoid this?

Regards,
Marco

Quote from: dkoppenhagen on January 06, 2017, 11:54:27 AM
I tried this. But it's the same result. So the "Data Type" in configuration section "General" is the type of the final values which is returned after my transformation script
runs on collected data? Or is this the type if the raw data which comes from my script?

As per my knowledge it is the type of the final value.

Quote from: dkoppenhagen on January 06, 2017, 11:54:27 AM
So the main problem is: I need a float for performance tab and a string for object overview.

So probably I have to leave it as it is and I change just the value description from "CPU" to "CPU (%)".

Since you need the same result with different data type, I suggest to leave it as float and change the description (you won't be able to show in performance graph if it is a string). Another solution is to duplicate the DCI and keep one as string and the other as float, the DCI of type string can be shown in overview and the float in the performance tab, but honestly if you have many nodes in my opinion is a waste of resources.

Regards,
Marco

Quote from: Victor Kirhenshtein on January 05, 2017, 07:29:30 PM
Hi,

there is no way to determine exact cause without access to agent log. In agent log you will find record started with text "LogWatch: Unable to open event log". Btw, if you can edit agent's config from console, you can enable file manager subagent and download agent's log file.

Best regards,
Victor

Unfortunately no access even for the config file (agent is configured to reply only to "Servers"), I'll ask for a copy of the log file to analyze it. Thank you.
Regards,

Marco

Quote from: dkoppenhagen on January 05, 2017, 10:48:34 AM
Hey guys,

I have a problem trying to transform a value which has been calculated from a script.
The script calculates the average CPU load over multiple CPUs.

It returns a float value (e.g. 41.34444444) which is the CPU load in percentage.
For displaying the data I want to transform this value so that it is displayed as "41.34444444 %".

Unfortunately the transformation doen't seem to work properly so in the object overview and in the view of last values I just see "41.34444444" (without "%").

Do you know what I configured wrong?

Thanks in advance.

Best regards,
  Danny

What is shown depends on the type of the DCI, so "41.34444444 %" is no more a floating point value, so you have to set it as a string and in your transformation script you have to do something like this:

string($1)." %"

in this case $1 is the value collected, change with the result of your calculation.

Regards,
Marco

Quote from: wraythe on January 05, 2017, 10:20:36 AM
Thanks Marco, what does the real function do?

Is used to manually convert the variable to a floating point. NXSL is loose typed programming language and the system should  automatically determine each variable type but I had problem to have it working correctly and that is the workaround I found.

Regards,
Marco

Hello!

I'v solved in this way:


if (real($1*8.0) > 10*1024*1024*1024.0)
return 0;
return $1*8;


in my case the cap is to 10Gib/s, change it according to your limit.

Regards,
Marco

I'm getting "EVENT LOG SUBSCRIBE FAILED" as return value for LogWatch.Parser.Status()

What can be the issue? I currently don't have access to the log of the agent 

Regards,
Marco

Are there any changes on RADIUS authentication starting from version 2.0.7? I'm no more able to validate users after upgrading from 2.0.5 to 2.0.7 

Regards,
Marco

Quote from: Marco Incalcaterra on November 13, 2016, 02:03:23 PM
Tested on server version 2.0.6 and agent version 2.0.5.

Any suggestions?

Anyone facing the same problem?

Regards,
Marco

As per subject, I have this problem when the agent is configured to listen from "Servers". If I configure it to listen from "MasterServers" everything works fine and I'm able to get the events.
Agents is sending properly the traps in both cases (I checked the log running manually with -D 6) but there is no track on the server when it is configured to listen for "Servers".

Tested on server version 2.0.6 and agent version 2.0.5.

Any suggestions?

Regards,
Marco

I had similar problem with a very old DLINK switch:

https://www.netxms.org/forum/configuration/snmp-monitoring-problem/

Sounds like a similar scenario.

Regards,
Marco