NetXMS Support Forum

Ответил в английской ветке: https://www.netxms.org/forum/configuration/server-and-agents-are-separated-by-nat/msg25297/#msg25297

Hi,

Geolocation is an immutable object, so you have to create new object when you need different location. NetObj and derived classes are also immutable, so you have to use NXCObjectModificationData class and session.modifyObject method. Working example for getting and setting geo location for node:


from org.netxms.base import GeoLocation

node = session.findObjectByName("netxms")
print node.getGeolocation()
md = NXCObjectModificationData(node.getObjectId())
md.setGeolocation(GeoLocation(11.0, 22.0))
session.modifyObject(md)


Best regards,
Victor

Currently no. We have a feature request for that.

Best regards,
Victor

Hi,

looks like in your setup you need agents connecting to server. In NetXMS default configuration is when server connects to agents. Reverse configuration called "agent tunnel". You have to configure server for accepting connections from agents: https://www.netxms.org/documentation/adminguide/server-management.html#server-configuration-for-agent-to-server-connection-tunnel-connection.

Best regards,
Victor

Hi,

you can only change it individually for each node. On agent side you have to add

ListenPort = nnn

to nxagentd.conf and on server side you should enter same port number in node properties Communication / Agent.

Best regards,
Victor

Hi,

replace <agent> tag with <core>. Also, it is not mandatory to use XMS format in policy - you can use same key=value format as in nxagentd.conf.

Best regards,
Victor

This version is very old. I definitely recommend upgrade to 2.2.12 first. We have fixed lot of issues, including internal locks and performance issues.

Best regards,
Victor

Hi,

this points to some internal locks being held for prolonged period. What version you are using, on what OS? Can you show event log for NetXMS server node?

Best regards,
Victor

Update: I likely found the reason for this error in numbering. Next release should show it correctly.

Best regards,
Victor

Hi,

you still need node to collect data from. On cluster level you can have only aggregated data. But you can define DCI on cluster level and associate it with cluster resource - that way it will be collected only on one node (which currently holds resource). If you need to do actual collection on other node (not cluster member) you can use "source node" option on DCI.

Best regards,
Victor

Проблема возможно связана с проблемой неработоспособности не UNICODE сборок, поскольку как я помню для OpenWRT собирается без UNICODE. У нас к сожалению нет никого кто разбирался бы нормально в OpenWRT сборках, поэтому функционал полуживой.

На локальном агенте загружен субагент portcheck? В nxagentd.conf должна быть запись

SubAgent = portcheck.nsm

Hi,

what is device OID and what driver is selected? If it is PROCURVE, please check if node has custom attributes .procurve.isModular and .procurve.slotSize.

Best regards,
Victor

For DCI data it is simple - set retention to 30 days, wait for housekeeper run (or run it manually), and set retention back.

Syslog is a bit trickier - if you just need to delete all old records you can do the same retention time change. But if you want to delete older records by certain criteria, you have to do it directly in database.

Best regards,
Victor

Somewhere mid-year I suppose, with first milestone release somewhere in February. We have lot of plans for 3.0 and many are still pending.

Best regards,
Victor