NetXMS Support Forum

Quote from: augusteDOUAMBA on March 06, 2024, 07:05:06 PMHello, I tried netxms on 500 node. It is a very powerful tool. Thank you and congratulations. I would like to deploy it on a wider scale. I don't have any planning tools on hand. Can the single server monitor 5000 nodes with 50000 dci?
If so, what should the characteristics of the server be?

Rough requirements are the same as mentioned above, 8-16GB RAM and 4 CPU cores, fast storage - ssd or nvme is recommended.
Attention should be paid to tuning of your database

Many of notification channel drivers depend on libcurl (and email sending was reworked to use libcurl somewhere in 4.x). How do you install the server, do you compile it yourself?

You can make a DCI with one of the following metrics: https://www.netxms.org/documentation/adminguide/service-monitoring.html#network-service-monitoring-using-dci

The DCI should be on a node that has agent running, e.g. on your NetXMS server node. But there's a way to overcome this - you can have this DCI on the node that you are monitoring and in properties of that DCI you can select another node in "Source node override" field.

In this case You'd need to create a Zone and set this worksation, which is on the same network as Synology NAS as proxy for that zone.

Could be firewall issue, some limits on server that prevent process from accepting connections. It was some time ago, so I hope it's resolved. If not, what's the output of
netstat -lptun
on your server?

Policy will be applied only to nodes, to which the template is applied (but should work in any sequence - either template was applied first and then policy was added, or otherwise).

On Windows systems default location for configuration policy files is
C:\Windows\System32\config\systemprofile\AppData\Local\nxagentd\config_ap
so you can watch this folder to check if policy gets there. Agent restart is needed so it can catch up the new configuration.


Also, on a side node, there are two ways how to monitor Windows Event logs - one is Windows Event Sync, when logs are sent to NetXMS server and stored there (so it requires bandwidth and DB space. The other is Log File Parser Policy - you can specify Windows Event Logs there in file name field, e.g. *Security

On your screenshot it's Event Monitor (which is updated in real-time). If you find this event in Event Log and double-click on it, you should have something like this:

No, in Users and groups all access rights are related to global things - e.g. ability to edit server actions - this is not related to any specific node. But things related to nodes and other objects - ability to modify configuration, edit comments and edit maintenance journal are set in the object tree - this way you can limit users to access only some nodes which is typically needed.

Right, looks like you are using Windows Event Log synchronization.

But how do details of NetXMS event "Win Event - Failed account log on" look in NetXMS event log?

"Disable usage of Modbus for all polls" sould be in Polling tab in properties of a node. But I thing there was some version of NetXMS where this checkbox was not present, so try upgrading to the latest.

As I remember, this information might be present in parameters of that event. If you open Event Log in NetXMS, find such event there and double-click it, a window should open with event details. Please share it here

Try connecting to agent port from your server with telnet:

telnet AGENT_IP 4700

it should establish connection.

The other thing to check is that your server IP is in agent configuration file in
MasterServers = IP_OF_YOUR_SERVER

Developers of NXSL may say in more details, but as far as I understand, the problem is that NXSL is loosely-typed language. So if you run this:
a = 33;
println(typeof(a));
a = "33";
println(typeof(a));

you'll see that when string contains only numbers, it's type is automatically converted to integer. Given that, using + both for addition and concatenation is problematic.

To add, since some time ago there are f-strings, e.g.:

value = 123;
println( f"The value is: {value}" );

so the need for concatenation might be not so high.

.. is already working for concatenation in v. 4.5. It's planned to have automatic conversion when migrating to 5.0 that would go through all scripts and change them.

In 5.0 there are compilation warning on use of deprecated functions, but these only present in script editor. The were some plans to check scripts everywhere in the system for deprecated functions, I will check later.

At some point in the future, when https://track.radensolutions.com/issue/NX-2322 will be implemented it would be possible to specify muliple OIDs and they will appear as variables in transformation script.

For now - either a script DCI that does all SNMP job and calculation and returns final result, or SNMP DCI that reads one OID and the rest being read in transformation script, e.g.:
totalMem = $1;

transport = $node->createSNMPTransport(); 
if (transport == null) return null;

freeMem = transport->getValue(".1.3.6.1.4.1.14823.2.2.1.1.1.11.1.4.1");
if (freeMem == null) return null;

return format(100UL * freeMem / totalMem, 1, 0);
The only bonus of the first approach is that you can perform recalculation for historical collected values.

Can you install netxms-dbg package and run netxms under gdb. It should be like:

gdb netxmsd
run

when it crashes, issue
bt
and share the output