NetXMS Support Forum

One possible reason is instances disappearing and re-appearing. If on one configuration poll instance is missing server will detele DCIs for that instance, and create new DCIs when instance appears again. You can check if this is the case by noting DCI IDs and check if they are changing.

Best regards,
Victor

External parameter runs on demand, when server requests it. So you effectively control it when setting DCI polling interval.

Best regards,
Victor

Немножко не так:


interfaces = GetNodeInterfaces($node);
foreach(i : interfaces)
{
   if (i->macAddr like "85:55:*")
      return true;
}
return false;


Насколько постоянные? Во время configuration poll сервер может перебирать разные версии протоколов.

Запрос для MS SQL:


SELECT TOP 200 event_id,event_code,event_name,severity,source_object_id,event_timestamp,message FROM alarm_events WHERE alarm_id=? ORDER BY event_timestamp DESC;



SELECT e.event_id,e.event_code,c.event_name,e.event_severity,e.event_source,e.event_timestamp,e.event_message
FROM event_log e,event_cfg c WHERE e.root_event_id=? AND c.event_code=e.event_code;


Простым образом сделать не получится. Через API можно получить объекты на карте, но отрисовку придется делать самостоятельно.

У меня скрипт отработал нормально (после замены sn на subnet во второй строке). Попробуйте включить дебаг уровня 6 на сервере и посмотреть что на стороне сервера происходит. И какая версия NetXMS? Я проверял на 2.1-М3.

Hi,

are you using instance discovery?

Best regards,
Victor

Hi,

you can define same DCIs on ESXi and set source node to management server. That way actual data collection will occur on management server but data will be shown on ESXi node.

Best regards,
Victor

Hi,

seems to be a bug in the code. What version you are using?

Best regards,
Victor

Hi,

seems that template was somehow removed from object structure but not deleted. Try to login with user system - most likely you will see that template in the root of object tree. Then you'll be able to delete it completely.

Best regards,
Victor

Hi,

please turn on agent crash dumps by adding to nxagentd.conf:

CreateCrashDumps = yes
DumpDirectory = some_path

and send crash dumps (if any) to [email protected]

Best regards,
Victor

Hi,

you have to add object access rights as well. They are assigned in object properties. You can also add your users to Admins group - it should have all object rights assignments already.

Best regards,
Victor

Hi,

you can use instance filter script to modify instance itself or just instance name. For example, if you instance is two numbers connected with dot you can use the following filter script:


instance = $1;
if ($1 ~= "[0-9]+\\.([0-9]+)")
{
   return %(true, instance, $1);
}
return true;


This script will keep instance as is (i.e. 4.2) but set instance name to second number (i.e. 2). You can then use {instance-name} in description to insert updated display name.

Best regards,
Victor

Hi,

alarms are grouped based on alarm key - if alarm with same key already exist when new alarm is generated new alarm will replace existing alarm and increase repeat count. Make sure you have meaningful alarm keys (or empty keys to disable alarm matching).

Best regards,
Victor