NetXMS Support Forum

Hi!

Yes, you specify top-level directory in --prefix, like usr. Make process will add bin, lib, etc. as needed.

Database schema version may or may not change between releases - it depends on actual requirement for changing it. Looking at change log, it seems that 1.1.7 was a minor release, mostly with bug fixes, so database schema was not changed.

Best regards,
Victor

Странно. Написал вот такой тестовый скрипт:


C:\Source\NetXMS\x64\debug>cat test13.nxsl
if ($1 == null)
        $1 = 1.234567;

println "floor(" . $1 . ") = " . floor($1);
println "ceil(" . $1 . ") = " . ceil($1);
println "round(" . $1 . ") = " . round($1);
println "round(" . $1 . ", 2) = " . round($1, 2);


Результаты:



C:\Source\NetXMS\x64\debug>nxscript test13.nxsl
NetXMS Scripting Host  Version 1.2.6
Copyright (c) 2005-2012 Victor Kirhenshtein

floor(1.234567) = 1.000000
ceil(1.234567) = 2.000000
round(1.234567) = 1.000000
round(1.234567, 2) = 1.230000

C:\Source\NetXMS\x64\debug>nxscript test13.nxsl 2.378
NetXMS Scripting Host  Version 1.2.6
Copyright (c) 2005-2012 Victor Kirhenshtein

floor(2.378) = 2.000000
ceil(2.378) = 3.000000
round(2.378) = 2.000000
round(2.378, 2) = 2.380000

C:\Source\NetXMS\x64\debug>nxscript test13.nxsl 2.98
NetXMS Scripting Host  Version 1.2.6
Copyright (c) 2005-2012 Victor Kirhenshtein

floor(2.98) = 2.000000
ceil(2.98) = 3.000000
round(2.98) = 3.000000
round(2.98, 2) = 2.980000

C:\Source\NetXMS\x64\debug>


Т.е. round(x, 2) работает как описано. По поводу 0 в конце: после floor/ceil/round, тип данных у результата остается float - поэтомы и цифры после запятой. Это правильно, если число будет использоваться в дальнейших вычислениях. Для красивого форматирования числа можно использовать функцию format (http://wiki.netxms.org/wiki/NXSL:format), отдельно или как обертку вокруг ceil/floor/round.

Пример использования format:


C:\Source\NetXMS\x64\debug>cat test14.nxsl
if ($1 == null)
        $1 = 1.234567;

println "format(" . $1 . ", 1, 2) = " . format($1, 1, 2);
println "format(floor(" . $1 . "), 1, 0) = " . format(floor($1), 1, 0);

C:\Source\NetXMS\x64\debug>


Результаты:


C:\Source\NetXMS\x64\debug>nxscript test14.nxsl
NetXMS Scripting Host  Version 1.2.6
Copyright (c) 2005-2012 Victor Kirhenshtein

format(1.234567, 1, 2) = 1.23
format(floor(1.234567), 1, 0) = 1

C:\Source\NetXMS\x64\debug>nxscript test14.nxsl 17.5672334
NetXMS Scripting Host  Version 1.2.6
Copyright (c) 2005-2012 Victor Kirhenshtein

format(17.5672334, 1, 2) = 17.57
format(floor(17.5672334), 1, 0) = 17

C:\Source\NetXMS\x64\debug>


Hi!

We are trying to find it. Can you describe what features you are using? Most important is to know about network discovery and scripts. Also, can it be that you have routers with very large routing tables (like full BGP table) monitored by NetXMS?

Best regards,
Victor

Hi!

Actually, event log tells you the same: that there are no description for the message. Looks like Snort don't register event source correctly. Quick search gives couple of people with same problem but no solution. I would suggest to reconfigure Snort to send it's log to text file instead of WIndows Event Log, and capture events from that text log file.

Best regards,
Victor

Hi!

Stack traces from different info files seems irrelevant to each other and don't give enough information. Can you please run server with debug level 6 (set LogFile parameter in netxmsd.conf to some file) and after next crash send me .info file together with log file. You can send them to my email to avoid posting sensitive information on the forum.

Best regards,
Victor

Hi!

That's strange. Seems that you are still running version 1.1.6. On what platform your NetXMS server is running? How do you do an upgrade?

Best regards,
Victor

Hi!

Check that user you are running agent under has access to log file specified in nxagentd.conf, and that path to the file exists.

Best regards,
Victor

Hi!

I'll try to reproduce it today. As a workaround, you may try to use odbc.ddr instead of mssql.ddr, but in that case you'll have to create database and db user manually.

Best regards,
Victor

Hi!

Would be interesting to see the results. There are no plans for dramatic changes in DB schema, so you are safe here

Best regards,
Victor

Hi!

Just committed changes necessary to run on RAP 2.0 RC2 - mostly changing dependencies to 3.8.0. It's strange, but in RC2 versions of eclipse packages changed to 3.8 from 3.9. After these changes, web ui seems to work fine. The only one strange issue I encounter so far is extremely large fonts in some places. I'll investigate it a bit later.

And I created new runtime platform from scratch. I encounter some strange issues in the past when trying to upgrade existing runtime platform, so to be on a safe side it's better to always create new one.

Best regards,
Victor

Do you shutdown it, or it just crashed because of no memory? Because log is the same. Don't wait for the crash - you have to run it for some time and shutdown correctly - otherwise valgrind will not be able to analyze address space of the process and find memory leaks.

Best regards,
Victor

Great. Now let it run for some time (watch that memory used by netxmsd grows, so memory leak is catch), then stop the server and send me the log again. Valgrind will record lost memory blocks only after process termination.

Best regards,
Victor

Normally it should not be like this. Sometimes valgrind slows down application significantly - if you have lot of nodes, initialization could take time. Try to wait for 5-10 minutes.

Best regards,
Victor

You can see that on the server. Open debug console and type "show sessions" command. It will show encryption protocol for each user session.

Best regards,
Victor

Hi!

Yes, Android console uses same communication protocol as PC console. If you turn on encryption, actual encryption algorithm is negotiated. If possible, AES-256 will be used, next option is AES-128, and last resort option is Blowfish. Login and password transmitted to the server after encryption setup. You can enable account block after number of unsuccessful login attempts by setting IntruderLockoutThreshold configuration parameters. Time of block determined by IntruderLockoutTime parameter (in minutes).

Best regards,
Victor