NetXMS Support Forum

Hi,

if IP addresses are readable from vendor MIB on Netonix and Ubiquity AirOS devices we can add a driver for them to create management interfaces correctly.

Best regards,
Victor

Hi,

default ordering is currently hard-coded. Feel free to add feature request for that.

Best regards,
Victor

Hi,

currently answer to both questions in "no". Feel free to register feature requests in our ticket system though For polling time change could be quite easy - server can measure and store polling time for last status poll, which can be made available through internal DCI. Having this value exposed you'll be able to do usual stuff like set thresholds and generate alarms.

Possible workaround cold be to create script DCI which will do SNMP request on any OID and measure time taken by that request.

Best regards,
Victor

Yes, %* formatter was added much layter then original topic was discussed. Now it is better to collect value as is and use formatters to display in K/M/G.

Best regards,
Victor

Hi,

most likely problem caused by trusted node check. You should either add current node (where script is being executed) to list of trusted nodes on node sw, or disable trusted nodes check completely by setting server configuration parameter CheckTrustedNodes to 0 and restart server.

Best regards,
Victor

Hi,

it was supposed to be fixed in 2.1-M2. Could you please give more details about your environment when 2.1-M2 is now working? Also, could you please try different browsers?

Best regards,
Victor

Just thought that if interface name poll works for you, so should configuration poll. How often you run configuration polls?

Best regards,
Victor

Hi,

I think option 1 is more feasible. There is Java API for reading DCI data, so all is needed is to find some Java library that can generate graphs in PNG format and write a simple web service around it. Do you know any good Java library for that? We could possibly use BIRT and built this functionality into web UI, but it will be quite heavy solution.

Best regards,
Victor

Hi,

#1 - seems like a bug, we will check this.

#2 - how exactly coordinates can be obtained? Is is possible to set geolocation from script, so for example if it can be obtained through SNMP it can be set in configuration poll hook. You can also use nxshell to run Python scripts for updating object coordinates from outside.

#3 - they just overlaps on map. We have to implement some sort of selector when there are multiple objects overlapping. I've added an issue for that: https://track.radensolutions.com/issue/NX-1162

Best regards,
Victor

No, you can only get value.

Best regards,
Victor

Hi,

yes, it's known bug in framework we are using (Eclipse RAP). We will try to find a workaround before stable release of 2.1.

Best regards,
Victor

Hi,

this is correct behavior - instance here is OID instance, which is determined by OID suffix (part after walk base). If you want to give other name to the instance, you can use instance filter to change it (and in DCI description you can use {instance-name} then). For example:

return %(true, $1, "New name for " . $1);

Best regards,
Victor

Hi,

you don't need instance discovery here. Script 2 is correct, you just have to place it into script library and use in script DCI. You can test it by selecting "Execute server script" on your switch node object, selecting library script or pasting script code, and executing.

Best regards,
Victor

Hi,

I've added to new indexes in 2.1-M3:

CREATE INDEX idx_syslog_source ON syslog(source_object_id);
CREATE INDEX idx_snmp_trap_log_oid ON snmp_trap_log(object_id);

Best regards,
Victor

$dci is global variable containing reference to current DCI object. You can find class description here: https://wiki.netxms.org/wiki/NXSL:DCI. Instance is it's attribute.

Transformation scripts in templates never executed - it is just a template for real DCIs.

It is not possible to disable DCI from script.

Best regards,
Victor