NetXMS Support Forum

Hi all!

We just published another patch release for 4.0 that fixes some newly found issues:

NX-2212 (Agent segmentation fault when accessing table Hardware.NetworkAdapters)
NX-2213 (Set event parameter from NXSL by index)

It also adds new parameter Oracle.Cursors.MaxPerSession to Oracle subagent.

Best regards,
Victor

Hi,

no, not cross compile - just if you have a choice of platform then Linux is easier to work with.

Best regards,
Victor

Hi,

most likely there was short period of time when system clock was set at 24.08.2033. Now server is not scheduling those DCIs to poll because current time is less than last value timestamp. Try to stop the server and run nxdbmgr check - it should delete incorrect records thus allowing normal data collection scheduling.

Best regards,
Victor

Hi,

you need Visual Studio 2017, and some precompiled 3rd party libraries: OpenSSL, libssh, libmosquitto, libcurl, database client libraries. You should add those libraries to include path and library path of Visual Studio. You will also need flex, bison, and perl in your path.
In general, setting up Windows build environment is much more complicated than Linux one, so if you have a choice, I'd really recommend going with Linux.

Best regards,
Victor

Hi,

you can set custom attribute on a node to current time from first event, and for the second event use script macro to read custom attribute value and return the difference with current time. For example:

1. Create script in script library called SetSignalLossTime:

$node->setCustomAttribute(".signalLossTime", time());


2. Create script in script library called GetSignalLossDuration:

t = $node->getCustomAttribute(".signalLossTime");
return (t != null) ? time() - t : 0;


3. Create server action of type script that will execute script SetSignalLossTime.

4. Add action from #3 to EPP rule where you generate alarm for signal loss.

5. In the message template for recovery event add %[GetSignalLossDuration] to insert duration of signal loss.

Best regards,
Victor

Hi all!

We just published another patch release for 4.0 which fixes two important issues:

NX-2210 (Network Service HTTP polling ports above 32767 generates incorrect Host header in HTTP request)
NX-2211 (Agent crash while processing request for System.ThreadCount)

Best regards,
Victor

Hi all!

We just published patch release for version 4.0 (4.0.2148). Changes since previous release:

- New attribute "lastCollectionTime" in NXSL class "DCI"
- Notification channel driver for Twilio
- Fixed server crash on receiving SNMP trap if configuration parameter SNMP.Traps.LogAll is set to true
- Fixed bug in automatic object binding
- Fixed bug in NXSL function FormatMetricPrefix
- Fixed bug in business service state calculation
- Fixed incorrect handling of manually set Y axis range on bar charts
- Fixed issues:
        NX-2153 (Add REST API endpoint for persistent storage)
        NX-2198 (Alternative data source for agent parameters Hardware.System.* on Linux)
        NX-2209 (cpuid.h is not supported on ARM macOS)

Best regards,
Victor

Hi, do you have core dump from crash? If not, could you run netxmsd under gdb until it crashes? If you can, make sure that you have installed the following packages:

netxms-server-dbg
netxms-dbdrv-mariadb-dbg

The run as following in terminal:

gdb netxmsd
On gdb prompt, enter
run -D1

When server crashes, you will get gdb prompt again. Enter command
bt

and share the output.

Best regards,
Victor

We just uploaded patched version of UI (version 4.0.2111).

Best regards,
Victor

Hi,

probably just ignore them, or set EPP filter to prevent generation alarms. Events will be generated only once for each DCI, so they should stop appearing when new templates will be applied to all old agents.

Best regards,
Victor

This actually looks like a bug. Can you provide configuration of your dashboard (you can export it into XML by right click on dashboard object and selecting "Export")?

Best regards,
Victor

Hi,

if you have DNS names correctly updated, then you can use DNS names of those devices for communications. NetXMS server will periodically (at configuration polls) check if IP address is the same. You can also force server to make those checks more often by changing configuration parameter Objects.Nodes.ResolveDNSToIPOnStatusPoll from "Never" to "Always" or "On failure".

If you have automatic network discovery then you can turn on parameter NetworkDiscovery.UseDNSNameForDiscoveredNodes - when set server will use DNS name instead of IP address for newly discovered nodes. You should have reverse DNS to be set up properly for this feature to work.

Best regards,
Victor

Could you enable SQL trace for some time (this will generate lot of records in log)? Then it will be possible to see all bindings for this query.

You can turn it on while server is running with the following commands in debug console:
debug sql on
debug db.query 9

and to turn back off:
debug sql off
debug db.query -1

Best regards,
Victor

Да, у нас с нескольких сторон есть такие запросы. Думаю сделаем в ближайшем будущем.

Hi,

yes, missing policy database error is related to SQLite issue. We will check today.

Best regards,
Victor