NetXMS Support Forum

The reason is that for some historical reasons of how communication with nxmc was build before web api ever existed, in some places several fields should be present in one request.

For snmpAuthName you also have to set snmpAuthPassword, snmpPrivPassword, snmpAuthMethod, snmpPrivMethod.

I've added comments in the documentation on this.

Try stopping Tomcat and deleting work catalog, then redeploying .war file again.

Try 4.0.2092 agent version.

If it's on Windows, try updated package netxms-server-4.0.2092-x64.exe

4.0 was just released. All business-service related functionality was added to the "old" nxmc.

Some brief explanation of new business services is in this text:

Business services that were present earlier were totally rewritten with tons of new functionality.

In a nutshell, Business Services is a tool for availability monitoring of logical services. Company email, web site, server farm, call center - all are examples of logical services. Or, these could be services your company is providing to others, e.g. if you are an ISP or MSP.

Moreover, the services can be combined together to define a "broader" logical service. For example, company DNS, web site and firewall all can be referred to as "Company Internet Services" and monitored for availability as a whole. So if the name server goes down then the "Company Internet Services" do not function properly as a whole. This feature can be used both for internal QA and external Service Level Agreement (SLA) monitoring.

Status of a business service depends on status of objects, DCI thresholds or script execution results. These things are called checks. Each business service can have multiple checks. Checks can be created manually or automatically with the help of auto bind scripts.

It's also possible to populate business services automatically – they are created on the basis of Business Service Prototype objects using instance discovery mechanism (similar to DCI instance discovery methods). Prototypes can have a number of predefined checks or scripts for automatic creation of checks.

Availability charts also got more flexibility. Now it's possible to get pie chart and detailed information about problems occurred for a business service, start time, end time and reason. Availability has some predefined time ranges or date selector for arbitrary date range.

New functionality provides more manager or business level information about your environment health and SLA compliance. Extensive scripting and discovery allows to automate creation and operation of business services. As the result, business services can give vital measurable information about your infrastructure work.

Permissions are configured in the object tree. Then are inheritable, so if something is configured on parent level, this goes down to children objects.

For dashboards two things are needed. First, there should be read access to the dashboard itself. Second - read access to all nodes that are used on the dashboard.
You can just give read access to these nodes in properties of these nodes. Or you can make a container under "Infrastructure services", put all needed nodes there and give read access to that container - that will be inherited down to nodes. Since you are not giving any access right to "Infrastructure services" container, that user won't see any nodes, he/she would only see Dashboards.

For 5% - could it be that EXT3 file system reserves about 5% of space for it's needs and actually does not allow to write files if it comes to that limit? E.g. some info here: https://linuxaria.com/pills/free-reserved-ext3-ext4-filesystems

About 16% threshold - thresholds are processed from top to bottom. If "process all thresholds" is unchecked, which is the default, thresholds are processes only until first one triggers. So if you have e.g. 4% value it will trigger 16% threshold and won't go any further. So the 16% should go to the bottom of the list.

Do you use any nxshell scripts, nxpush, web api or anything that would connect to netxms server for automation purposes?

Approach with custom attributes is quite OK - that way scripts can be simplified.

You can also run this script on any node (and you can even make an Object Tool that would execute this scrip):
for (n: GetAllNodes())
{
  if (!n->isParent(FindObject("Infrastructure services")))
  {
    println(n->name);
  }
}

And there's a good place - Tools->Find Object. Switch to query tab there and put the following query:
type == NODE and !$node->isParent(FindObject("Infrastructure services"))

It will find all nodes that are not present under "Infrastructure services" and it's an interactive list - you can right-click nodes and do some operations with them.

Not yet implemented, you can follow this issue to see when it will be: https://track.radensolutions.com/issue/NX-2114

For now you can create a chassis, give it a custom image and place in the rack. The only downside is that it will have Unknown status.

What data type do you use for FileSystem.Free({instance})? Try using "Unsigned Integer 64-bit" there as if you have a big HDD, the value can be very big as it's in bytes.

If it's still not working - how exactly the DCI is behaving? Does it goes into error?

So yes, if you manually configure custom attribute with same name on a child node, this overrides the value which is being inherited from a parent container.

User interface was a bit reworked, so in version 4.0 that is about to be released soon it should be easier to understand what custom attribute is defined locally and what is inherited.

The logic with custom attribute is the following. Say we have 3 levels of hierarchy:

|__A
     |__B
          |__C

Say we define an inheritable custom attribute on A. It will populate to B and C.

We can then redefine (override) the same custom attribute on B (by editing it) and give it a new value. That new value will also be populated to node C.

The tricky part is that UI is quite confusing here. A custom attribute that comes inherited from a parent node is displayed on gray background in the list.
When redefined, it is displayed on white. When redefined, it still displays a node in "Inherited from" column, but it has meaning "if you delete redefined custom attribute here, a value will come from that node".

We will discuss internally how the UI could be improved on this.

Ну нет, какого-либо стандарта xml файлов между системами мониторинга нет. В NetXMS есть экспорт большинства вещей из конфигурации в XML - шаблонов, событий, правил EPP. Но Zabbix это совершенно другая система построенная на других идеях, так что эти данные туда не заимпортить.

Исторические данные хранятся в базе, по отдельной таблице на каждую ноду. Формат там простой, на каком-нибудь питоне наверноем можно накодить конвертор.

Можно полюбопытствовать о причинах перехода на Zabbix?

What is the exact version of NetXMS?