NetXMS Support Forum

With NXSL it is possible to generate an output with "println" to the client.
Is there a possibility to create an input-window from an nxsl-script where an input can be requested with an individual description (value=readln("Please insert a value:")). Something like You have implemented with objecttools "InputFileds" but even generated as a result of some script-calculation.
Best regards
Georg

There is a perl script called mysqltuner that helps very much identifying unused space within the tables that will lead to large files in the database.
Especially after changing some RetentionTimers and deleting some records via housekeeper you will see spaces there where you can shrink databasesize.
Sometimes it is good identifying unused tabels via nxdbmgr.

Hello Filipp,
I followed Your advice and looked at the length of the array in called function and in calling function. Value was the same both times e.g. 4.
Reducing and testing code via calling the script as serverscript from a device. This scriptlib has the subs and a "main"-function that calls the subs. With reduced code resulting string appears in the calling-function. After introducing some more calculation in called function the problem appears once more with the following results:
NT_lib_InstDisc::int_InstDisc_01 called with: ##---#nsnsnsn bhohofewho
NT_lib_Parsers::Parse_Interface_Alias_Prefix called with: ##---#nsnsnsn
Parse_Interface_Alias_Prefix IN default:13 ##---#nsnsnsn
NT_lib_Parsers::Parse_Interface_Alias_Prefix returns Values (hashash;has2hash;hasparams;param;;arraylength):1;1;1;---;;4;
Error 15 in line 37: Unknown object's attribute

Some times automatic server-restart followed.

Putting a trace command directly in front of building  retval changed situation and function worked as expected.

WITH LINE 135 NOT COMMENTED OUT:
NT_lib_InstDisc::int_InstDisc_01 called with: ##---#nsnsnsn bhohofewho
NT_lib_Parsers::Parse_Interface_Alias_Prefix called with: ##---#nsnsnsn
Parse_Interface_Alias_Prefix IN default:13 ##---#nsnsnsn
NT_lib_Parsers::Parse_Interface_Alias_Prefix param:---
NT_lib_Parsers::Parse_Interface_Alias_Prefix returns Values (hashash;has2hash;hasparams;param;;arraylength):1;1;1;---;;4;
In int_InstDisc_01 Return Values from Parse_Interface_Alias_Prefix: 1;1;1;---;;4;

As of this You will find complete code I wrote as attachment, because I don't know where my error is.

Best regards

Georg

I'm not genius in script writing. As of this I do not know what I'm doing wrong.
Using Netxms 4.0-2227 calling a script from scriptlibrary out of Instance discovery filter I'm doing some analyzing of alias-string.
>>> 2022.04.15 22:38:50.538 *I* [nxsl.trace         ] NT_lib_InstDisc::int_InstDisc_01 called with: ##---#hjipeffebguo Fa0/4 (##---#hjipeffebguo)
Within the sub int_InstDisc_01 an other sub is called that should return an array consisting of three values and a string with three characters.
>>> 2022.04.15 22:38:50.538 *I* [nxsl.trace         ] NT_lib_Parsers::Parse_Interface_Alias_Prefix called with: ##---#hjipeffebguo
Using trace function I can figure out that the array will be constructed the right way within called sub Parse_Interface_Alias_Prefix.
>>> 2022.04.15 22:38:50.538 *I* [nxsl.trace         ] NT_lib_Parsers::Parse_Interface_Alias_Prefix returns Values (hashash has2hash hasparams param ):1;1;1;---;
But printing the resulting array within int_InstDisc_01 I only can identify the three values but there is no string to find
>>> 2022.04.15 22:38:50.538 *I* [nxsl.trace         ] In int_InstDisc_01 Return Values from Parse_Interface_Alias_Prefix: 1;1;1;;
Moving the three character-string from the last place in the array a little bit forward within sub Parse_Interface_Alias_Prefix just gave me analogous result without the string in returned array.

Is there a restriction concerning mixing of different types for the return array of a sub?

For production environment I have the two scripts in different libs. But I build up a test system where I put them in one library to reduce complexity.


Script snipplets:
sub int_InstDisc_01(ifalias,ifname)
{ trace(0,"NT_lib_InstDisc::int_InstDisc_01 called with: ".ifalias." ".ifname);
   
  inst_disc = false;
  array res1;

  res1 = Parse_Interface_Alias_Prefix(ifalias);
  trace(0,"In int_InstDisc_01 Return Values from Parse_Interface_Alias_Prefix: ".ArrayToString(res1,";").";");
.....
}

sub Parse_Interface_Alias_Prefix(ialias)
{ trace(0,"NT_lib_Parsers::Parse_Interface_Alias_Prefix called with: ".ialias);
 
  array retval;
....
  retval = %(hashash,has2hash,hasparams,param);
  trace(0,"NT_lib_Parsers::Parse_Interface_Alias_Prefix returns Values (hashash has2hash hasparams param ):".ArrayToString(retval,";").";");
  return(retval);
} 


Since update to version 4.0 (include nxdbmgr upgrade) there will be the messages shown in the attachment within consolewindow.
My environment is debian with mariadb installed.
Currently netxms is compiled from source with --with-mariadb --with-mariadb-compat-headers.
Best regards and thanks for Your software
Georg

Hello together,
I compiled netxms-server on debian from source.
When starting netxmsd I got the following problem:
1.) Starting with parameter -c /etc/netxmsd.conf I'm always gettng an error "Error loading config file"
2.) Starting without this parameter worked.
Configfile is a link within /etc-directory. Thus same file is used when calling with or without -c option.
Other parameters like -d or -D are working.
Best regards
Georg

Hello together,

defining a gauge for a dashboard it is possible to set values for different colorzones. Within properties form they will be defined by hand.
Is there a possibility to configure a calculation of these values from the values of an other DCI?
At our environmental monitoring we have defined individual warninglimits for some sensors. These limits are read out with a DCI and it would be great coloring the actual values regarding the individual warninglimits.

Best regards

Georg

I did it within a Hook-script while configuration poll based on a prefix in the interface interface alias e.g. description.
For example ## at the beginning of the interfacedescription tells me to set expected state to "UP".
Second using this prefix within instance dicovery (instance discovery filter script) You can automatically generate DCIs for these interfaces.

Solution with editing the dci does not work for templates! After "save and apply" the dci within the template does not has the advice "This DCI was added by instance ...." but within the component the template dci is associated the advice appears. Editing this dci within the component will lead to the advice within the template once more.

After editing the dci (without changing it) only doing "apply and Save" the problem for the edited dci has disappeared.

Hello together,

I did a strange upgrade from version 2.0.5 with old mysql-database on openesuse to new mariadb 10.3 and netxms version 3.4.232 (with netxms dbupgrade and backgroundupgrade) on debian.

Using the GUI (via Tomcat and native gui on windows) first I was shoked, because from datacollection-configuration window I cannot display the linechart from a dci. Every time I'm trying this I will get "Cannot get value for DCI[0]:"<Name der DCI>" (Invalid object ID)".
But displaying last values from an object will work properly and within the "lastValues"-Window using linechart for a dci will display graph properly too. Same for normal DCi an table DCI. As I can see within lastvalues window datacollection is working regulary.
Editing a DCI from "Datacollection Window" I allways wil get the follwoing advice "This DCI was added by instance discovery Alllocal changes can be overwritten at any moment" but these DCIs are not edited with instance discovery!

Can anyone help me?

Best regards

Georg

Hello Victor,

thanks for Your fast response!

While preparing the download of the images for You I realized that there was a problem in display the copy of my file from my new server on windows too. I have seen, that there was a difference in bytecount  between the original file on my old netxmsserver and that on my new server.
I did the transfer via filezilla initiated from the new server one more time and got the same results (concerning bytecount differences). Filezilla was configured for automatic detection of transfertype (configured as standard). When configuring filezilla explecitly for binary transfertype the file is transfered correctly and has the original size.
This is a behavour I have never seen before with filezilla.

The images transfered in binary type via filezilla all work fine with the new implementation of Your image library tool now. For administration the responsetimes of the tool are really great now even for large images.

Sorry for posting the problem that was generated via an other tool!

Best regards

Georg

As a consequence to currupted display of the image. Object that are placed in a map will be misplaced to the backgroundimage.
Attached the same part of the building with Netxms 2.0.5 and Netxms 3.2

Hello Victor,

thanks a lot for Your work providing netxms to the community!

I like the principles of your new implementation of image library, but I have some problmes to use it. First I have to tell You three things:
1.) I'm using images that are imported with version 2.0.5 of netxms-software. These images are copied from old imagelibrary to new system.
2.) I'm using source-code which is compiled on debian 10.
3.) As GUI I'm using windows 10-Console.

My problems are:
1.) Display of some images is not possible?
2.) Display of some images is broken, incomplet ore disrupted
This happens within imagelibrary and on the maps that are using these images.

Are there any needs of other packages to be installed with the new implementation of images?

Best regards

Georg

Hello Victor,

OK.
Thanks for the information!

Thanks for your software too! It helps me a lot on monitoring our network and the services we provide at networklayer.

Best regards

Georg