NetXMS Support Forum

Right, currently the installed files are in several other places.

Please add --prefix=/opt/netxms to ./configure command, run ./configure ...., make and make install and send the contents of /opt/netxms folder


UPD: or if you have not done the above yet, please just send the contents of the following folders from your system.
/usr/local/bin
/usr/local/lib

The menu should be available on the login page.

The thing is, that empty custom attribute is converted into integer and turns into 0.
Not sure when there were changes in this. But current behavior is actually correct one - if custom attribute exists, but is empty, we treat it as existing and do not apply the default value for threshold. For numeric threshold empty value is converted into 0.

We also need exact binaries how they got compiled on your system. Please provide an archive of netxms/bin and netxms/lib folders.

Please proceed with netxmsd installation from packages: https://www.netxms.org/documentation/adminguide/installation.html#installing-on-debian-or-ubuntu

what about netxmsd service? what is the output of
systemctl status netxmsd
?

As for max_locks_per_transaction - I would first check general postgres parameters, first of all probably shared_buffer.
Out of the box Postgres settings are very conservative.
See e.g. https://www.percona.com/blog/2018/08/31/tuning-postgresql-database-parameters-to-optimize-performance/

Can you connect to postgres and show output of these queries:

EXPLAIN ANALYZE SELECT max(event_id) FROM event_log

EXPLAIN ANALYZE SELECT max(alarm_id) FROM alarms

Did you initialize timescale DB with nxdbmgr init prior to importing data into it?
Can you set debuglevel = 9, run the server for 5-10 minutes and show the log?
Alternative approach would watch from the database side on what queries it's getting stuck.

Is your tomcat on the same machine where netxms server is running?
Is netxms server process (netxmsd) running?

Please run
ulimit -S -c unlimited
this will enable core dump creation.
Then run
nxdbmgr init
again. When it crashes, a file should appear in current folder named core.xxxx. Please attach this file here.

As per some post on stackoverflow, mssql is able to get into a deadlock on simultaneous insert and delete due to some things related to indexes.

Some changes will be introduced into db writing mechanism in coming patch release. Currently, when housekeeping is happening, there are two separate queues - one for housekeeping tasks and another for new data inserts.
In the coming version a running housekeeping query will pause insert queries. After housekeeper query completes, it will check the length of current insert queue, if it's above Housekeeper.Throttle.HighWatermark setting housekeeper will pause (until reaching LowWatermark), otherwise another housekeeper query would run.
As a result this should fix the deadlocking issue.

Currently the only way to make this work is to create a separate node for each slot. In configuration for each node you have to specify corresponding community string.
To avoid issues with one and the same IP address for several nodes set checkbox "This is address of remote management node" for all these nodes. And disable configuration polls for these nodes.

If you have a big network of such devices, you can contact Raden Solutions, they can develop a custom driver for such devices within sponsored development.

Do you mean exactly "status map" dashboard element, or "Alarm viewer"?
Currently there is no filtering for both of them. But please describe in more details what do you want to achieve? Do you want to show status of a node as green, when alarm is acknowledged or something else?

На какой именно?