NetXMS Support Forum

In NetXMS the management console keeps all objects (containers, nodes, etc) in memory with their statuses. When object information changes on server, it sends information about this to connected consoles. If there are a lot of changes, but communication to a console is slow, it could happen that server won't be managing to send the updates. In this case server send a message to the console to display that warning.
Currently there is no healing mechanism, so once this message has appeared one would need to restart the console.
In 3.7, which is hopefully coming in next few days communication to console has been improved, so it might become better with this.

Not sure. At hand I have Centos 7.8, it has gcc version 4.8.5 20150623 (Red Hat 4.8.5-44) (GCC) and ./configure in 3.3 runs fine on it.

also not sure why you are specifying openssl, curl and pcre locations - are they not available from packages?

This might be the reason:
/usr/bin/ld: error while loading shared libraries: libbfd-2.20.51.0.2-5.48.el6_10.1.so: cannot open shared object file: No such file or directory

this file seems to be in binutils package.

Can you share your config.log?

Let's try to get some more debug information. Can you please run the following command in the command line on the server:

nxadm -c "debug client.* 6"
This will increase debug level for log messages related to communication with the console.
Then try to open IP neighbors window and wait for the error.
When the error shows, turn off the additional debugging
nxadm -c "debug client.* -1"
and send us server log file

Are you sure that nxagentd.conf portion is exactly the same that the agent is running with? Was agent restarted to work with actual config?
For me "RunExternal (shell exec): worker thread created" only appears when ExternalParameterShellExec is used (by the way this soon gonna be reworked and all actions and external parameters will be lauched as ShellExec).

I've tried with agent 3.6.300. Here's my config:
ExternalParameterShellExec=Get-ScheduledTask(*):powershell.exe -NoProfile -Command "$$Parameter = \"$2\"; Write-Host (Get-ScheduledTask -TaskName \"$1\").$$Parameter"

And that's the log:

2020.12.16 03:41:40.720 *D* [comm.cs.50         ] Requesting parameter "Get-ScheduledTask("Test Task", "State")"
2020.12.16 03:41:40.720 *D* [comm.cs.50         ] H_ExternalParameter called for "Get-ScheduledTask("Test Task", "State")" "Spowershell.exe -NoProfile -Command "$$Parameter = \"$2\"; Write-Host (Get-ScheduledTask -TaskName \"$1\").$$Parameter""
2020.12.16 03:41:40.720 *D* [exec               ] RunExternal called for "Get-ScheduledTask("Test Task", "State")" "Spowershell.exe -NoProfile -Command "$$Parameter = \"$2\"; Write-Host (Get-ScheduledTask -TaskName \"$1\").$$Parameter""
2020.12.16 03:41:40.720 *D* [exec               ] RunExternal: command line is "powershell.exe -NoProfile -Command "$Parameter = \"State\"; Write-Host (Get-ScheduledTask -TaskName \"Test Task\").$Parameter""
2020.12.16 03:41:40.720 *D* [exec               ] RunExternal (shell exec): worker thread created

Ethernet/IP is a protocol used on some industrial controllers, this message informs that netxms was not able to connect via this protocol.

PDH error is related to windows performance counters. On configuration poll netxms is trying to read the list of performance counters. If this was successful, when adding a DCI of type "Windows performance counters" under "Select" button there will be a list of available counters.
In node properties -> Polling it's possible to turn this off - "Disable reading of Windows performance counters metadata" checkbox.

Here's an old post that mentions same PDH error:
https://www.netxms.org/forum/general-support/problem-with-windows-performance-counters/

А как именно вы пытаетесь добавлять MIBы?

Во этих MIB файлах есть зависимость от SNMPv2-SMI.txt
Он должен был проинсталлироваться вместе с NetXMS в директорию mibs, туда же нужно добавлять ваши mib файлы и запустить mib compiler.

Did you open User Manager recently? From the same netxms console, or from another machine?
If you open Tools-> Server console, what is the output of
show sessions
command there?

Does this error stays, if you close netxms management console and start it again?

Most probably it will go away if you restart netxms server process.

What's the output of
show status where variable_name = "innodb_page_size";

I've just tried to install NetXMS 3.5 from packages on Debian 10 and upgrade to 3.6 went successfully. Could it be that your NetXMS database was created some time ago and on older MariaDB?

What is the value of innodb_page_size on your system?

Несколько путей:

а) в правиле в EPP в filtering script:
return ( (strftime("%H", time()) >= 8) and (strftime("%H", time()) < 19) );

Таким образом все правило в EPP не будет срабатывать ночью. Алармы также не будут создаваться. Если нужны алармы, то тогда нужно два правила в EPP.

б) не запускать action, отсылающий сообщение напрямую, а запускать action, вызывающий NXSL скрипт. В этом скрипте сделать проверку по времени и оттуда запускать отсылку сообщения:
if ( (strftime("%H", time()) >= 8) and (strftime("%H", time()) < 19) )
{
    SendNotification("Telegram", "Alice; Bob", "", "My message text");
}

Можно сделать Script DCI для сбора данных. Это может быть один скрипт, в который в качестве параметра передается OID.

В этом скрипте можно делать несколько попыток опроса с задежкой. Если все равно не получилось получить данные, можно сделать abort - в этом случае будет data collection erorr, но DCI не перейдет в unsupported.

Configuration poll is once per hour by default. You can try setting debug level 6 on one agent, restarting it and running configuration poll, in this case we will see at what moment this error is happening.
On what OS the agent is running?

Recent changes in NetXMS added fields to nodes table. Looks like we are now hitting page size limitation: https://mariadb.com/kb/en/innodb-row-formats-overview/#maximum-row-size
Increasing innodb_page_size might help.