NetXMS Support Forum

gdb should print out stack trace.

Sample session:
AirAlk:~() $ gdb ./crashapp
GNU gdb 6.3.50-20050815 (Apple version gdb-1820) (Sat Jun 16 02:40:11 UTC 2012)
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "x86_64-apple-darwin"...Reading symbols for shared libraries .. done

(gdb) run
Starting program: /Users/alk/crashapp
Reading symbols for shared libraries +.......................... done

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000000
0x0000000100000f93 in main (argc=1, argv=0x7fff5fbff8d0) at crash.c:7
7   *p = 123;
(gdb) bt full
#0  0x0000000100000f93 in main (argc=1, argv=0x7fff5fbff8d0) at crash.c:7
p = 0x0
(gdb)


I'm interested in part starting with "Program received signal..."

Hello.

Please provide backtrace of the crash. There are two options:

1) If "core" file is generated, get stack trace directly from core:gdb /opt/netxms/bin/netxmsd core"
then inside gdb:bt[/bt]

2) Start netxmsd under gdb:
[code]gdb /opt/netxms/bin/netxmsd

Then on GDB prompt, type (everything after run is parameters for netxmsd):run -c /opt/netxms/etc/netxmsd.conf -D0

When server crashed, type "bt" on gdb prompt to produce stack trace.

Quote from: Dani@M3T on May 19, 2014, 02:30:58 PM
I just upgraded my NetXMS Server from 2.1.13 (linux 64bit installed from sources). But the server crashes directly after start.

Note about Mikrotik support — it's already implemented and will be included in upcoming 1.2.14

Do you need to trigger on any change, or just some predefined list of software products?

Скорее всего проблема связана с тем, что в транзакции пересоздается большое количество таблиц (tdata_*), а PG делает при этом exclusive lock до конца транзакции. При большом количестве устройств или специфической настройке PG, локов может не хватать. Попробуйте увеличить значение max_locks_per_transaction в postgresql.conf (нужно только на время апгрейда).

Вопрос — сколько у вас устройств в системе?

Right now there are no proper way to do that. It's probably possible to make transformation script for the table, which will iterate over all rows, compare with previous value and and post events directly in circumvention of the threshold mechanism, but it will be quite a hack.
Right now I'd write a standalone script or application, which process records from tdata_* tables and then post events, either using "nxevent" command line tool or via Java/Jython API.

You are trying to call static method — code should be like this:

for node in [o for o in s.getAllObjects() if isinstance(o, objects.Node)]:
point = s.findConnectionPoint(node.ZONE0, node.getPrimaryIP())
print node.getObjectName(), point


And you don't need to import org.netxms.client, it's done automatically by nxshell.
from org.netxms.client import *
from org.netxms.api.client import *

Попробуйте такой запрос:SELECT n.primary_ip,o.name FROM nodes n, object_properties o WHERE n.id=o.object_id AND o.is_deleted=0

Можно сделать например так: File.Size(/path/to/backup_%d_%m_%Y) — сегодня вернет размер файла /path/to/backup_17_04_2014

Агента в планах пока нет, но вы можете сделать feature request тут: https://www.radensolutions.com/chiliproject/projects/netxms/, и мы его учтем.

Рестарт агента помогает?

No idea, to be honest. I'd try to capture traffic with tcpdump when nxdbmgr connects (successfully) and compare it to traffic when server connects (with error)

Quote from: jeffreyz on March 28, 2014, 05:40:52 PM
Do you have any other suggestions?  The data is not very important since this is for a server that I am only using for testing.  Plus it is on a virtual machine so I am willing to try anything - I can always revert to a snapshot if things go south. But I do need to get it going somehow.

Check this presentation: http://www.slideshare.net/xzilla/the-essentialpostgresconf

SQLite is mostly for development or really small installations.
All other DB engines are production ready. In most installations we use PostgreSQL, with proper tuning it works pretty good — but decision is usually based on existing infrastructure or DBA experience.
Currently we fully support DB2, Informix, MSSQL, MySQL, Oracle, and PostgreSQL.

Quote from: jeffreyz on March 23, 2014, 08:15:13 PM
I upgraded from 1.2.12 to 1.2.13 and now I cannot start the server because the server cannot log into the database.  Here is the output from the server log file (DebugLevel = 9):

1. Try to change "localhost" to "127.0.0.1"
2. Please show your pg_hba.conf

Encryption is used by default, however it's a subject to supported ciphers and negotiation during initial handshake — with plain text fallback. To ensure that network traffic is always encrypted (or connection is rejected), add "RequireEncryption=yes" to nxagentd.conf.