NetXMS Support Forum

Зарегистрировал как баг (https://www.radensolutions.com/chiliproject/issues/373), посмотрю попозже.

Добрый день!

Это баг просмотра настроенных действий, само действие должно работать правильно.

Looks like same issue: https://www.netxms.org/forum/general-support/segfault-2648. Let me know if you need compiled binaries.

Best regards,
Victor

Hi!

Was this SNMP or agent table? Looks very similar to https://www.netxms.org/forum/configuration/netxms-server-crash-1-2-9.

If you built server from sources, try to replace line 389 in file src/server/core/dctable.cpp which currently should looks like


if (_tcslen(s) < MAX_DB_STRING)


with


if ((s == NULL) || (_tcslen(s) < MAX_DB_STRING))


and rebuild server.

Best regards,
Victor

Hi!

Currently it's not possible to add a label to Y axis. I've registered it as feature request (https://www.radensolutions.com/chiliproject/issues/372).

You can transform uptime to days using transformation script like this:


return $1 / 86400;


You may need to change DCI data type to floating point number if you want to see fractional values.

Best regards,
Victor

Hi!

Good news is good

I've found possible reason for a crash. Are you using SNMP tables? I can build a patched binary for you so you will be able to test it.

Best regards,
Victor

Попробуйте поставить кодировку CP-1251 в netxmsd.conf. Если не требуются еще какие-то языки, то должно работать без проблем.

Hi!

You can create another DCI for System.CPU.Count parameter and read it's value in transformation script by calling GetDCIValueByName($node, "System.CPU.Count"). Full script could looks like


return $1 / GetDCIValueByName($node, "System.CPU.Count");


Best regards,
Victor

Думаю что можно будет. Добавил в feature requests: https://www.radensolutions.com/chiliproject/issues/368

А можете stack trace прислать (команда bt в отладчике после получения SIGSEGV)?

Sorry, not yet. At some point I completely forgot about it I've added a bugtracker issue (https://www.radensolutions.com/chiliproject/issues/367) so it will not be forgotten next time.

Best regards,
Victor

Hi!

Yes, you can create node object and then define DCI with data really being collected from another node. You should set "proxy node" property in your DCI to the node where data collection really occurs. For example, you have server node (MGMT) and monitored node (TARGET). You configure PING subagent on node MGMT. Then create DCI for ping subagent on node TARGET with proxy set to MGMT. It will collect data from MGMT, but it will be associated with TARGET.
You have to add node where you create proxied DCI to proxy node's "trusted nodes" list, or turn off trusted nodes checking by setting server configuration variable CheckTrustedNodes to 0. A bit more detailed explanation of trusted nodes concept is here: http://wiki.netxms.org/wiki/SG:Security_Issues.

Best regards,
Victor

You cannot measure latency or packet loss with method 1. If you need it, you should use method 2.

Best regards,
Victor

Hi!

You can use instance discovery for that. For example, to monitor input bytes on all interfaces, create DCI (either directly on node or in template) with SNMP OID set to

.1.3.6.1.2.1.2.2.1.10.{instance}

and on "Instance Discovery" page choose SNMP Walk - Values as discovery method and enter .1.3.6.1.2.1.2.2.1.1 as SNMP OID.

Best regards,
Victor

Hi!

Rule should looks like this:


<rule>
   <match>^\[INFO \]\[SERVER\] (.*)</match>
   <event params="1">event code here</event>
</rule>


Best regards,
Victor