NetXMS Support Forum

Hi!

Didn't you miss some alarms? It can happen in this situation that node down alarm didn't get terminated automatically, and you have to terminate it manually. If this is not the case, than it's definitely a strange bug.

Best regards,
Victor

Version 1.0.5 just released, where I have changed licensing for most libraries to LGPL. Current licensing scheme is as follows:

NetXMS components:

libnetxms       LGPL
libnxcl         LGPL
libnxdb         LGPL
libnxlp         LGPL
libnxmap        LGPL
libnxmb         GPL
libnxsl         LGPL
libnxsnmp       LGPL
libnxsrv        LGPL
nxagentd        GPL
nxalarm         GPL
nxav            GPL
nxcon           GPL
nxcptest        GPL
nxevent         GPL
nxflowd         GPL
nxhttpd         GPL
nxlexer         GPL
nxlptest        GPL
nxmibc          GPL
nxnotify        GPL
nxpush          GPL
nxscript        GPL
nxsms           GPL
nxsnmpget       GPL
nxsnmpset       GPL
nxsnmpwalk      GPL
nxuilib         GPL

all subagents   GPL
all DB drivers  GPL
all SMS drivers GPL


Third party components bundled with NetXMS:

freetype        BSD-like
libexpat        BSD-like
libgd           BSD-like
libipfix        LGPL
libjpeg         BSD-like
libpng          BSD-like
scilexer        BSD-like
sqlite          BSD-like
zlib            BSD-like

Best regards,
Victor

Hi all!

NetXMS version 1.0.5 is out. Changes since 1.0.4:

- New NXSL operators: += -= *= /= .= %= &= |= ^=
- New NXSL functions: index, rindex
- MIB compiler improved
- New MIBs added to default distribution:
   ASTERISK-MIB, DIGIUM-MIB, NetWare-Server-MIB, NETWORK-APPLIANCE-MIB
- Windows console: fixed problem with saving Object Browser and maps in
 desktop layout
- Fixed issues: #279, #303, #304, #305

Best regards,
Victor

Fixed in 1.0.5. Thanks for reporting!

Best regards,
Victor

I have added missing files to dist in 1.0.5. Also, resources for most of the projects may not load until message files compiled to binary form - try to compile messages.mc first.

Best regards,
Victor

Добрый день!
А можно прислать скрипт disksp.sh? Попробую у себя погонять. И еще вопрос - агент под рутом запущен?

Hi!

You are using installation wizard, right? What user you are using as DBA account? Does it have rights to create new database? If you not using Windows authentication, do you enable SQL Server authentication?

Best regards,
Victor

Yes, you have to configure traps to events matching manually for each trap. Also, if you will go that way and create trap configuration for Juniper traps, it will be nice if you can export trap configuration and publish it here.

Best regards,
Victor

Hi!

You have two options:

1. Create an action of type "execute command on management server", entering your command in "command" field like following:


c:\\Rcmd \\\\exchange1 osql -S exchange1 -U sa -P ourpassword -Q "insert into VGSMDB..loglog (destination, content) values ('Majors','Server %n is down')"


Note that all backslashes are doubled, because backslash is an escape character in action command definition, and you have to use \\ to get single backslash in command.

2. Modify you .bat file to accept server name as argument:


@echo off

c:\Rcmd \\exchange1 osql -S exchange1 -U sa -P ourpassword -Q "insert into VGSMDB..loglog (destination, content) values ('Majors','Server %1 is down')"

pause


and define action to call it, with "command" field set to


my_script.bat %n



In both cases, just add this new action to appropriate rule in event processing policy.

Hope this helps!

Best regards,
Victor

Pohoze na issue 305 iz bugtrackera, no ja esche poizuchaju dump. Esli prichina ta-ze, to v 1.0.5 eto ispravleno, segodnja ili zavtra ja sdelaju build.

The current scheme is "for historical reasons". Now I have much better understanding of how it should work, but it's a major change in server's architecture. I completely agree that physical interfaces and IP addresses must be separate entities. I plan to make this change in 1.1.x branch (along with many other required changes).

Best regards,
Victor

Hi!

It's not possible to just graph data in different units. However, you can convert bytes to bits at collection time using transformation script, and then work everywhere with bits.

Best regards,
Victor

Hi!

It's impossible to extract part of the varbind's OID as event's parameter. However, it's possible to extract varbind's value. If trap contains some value identifying BGP peer, then it can be extracted. I suppose that BGP trap can contain either .1.3.6.1.2.1.15.3.1.7.<peer_ip_here> (bgpPeerRemoteAddr) or .1.3.6.1.2.1.15.3.1.5.<peer_ip_here> (bgpPeerLocalAddr). In that case, you can add varbind matching in trap configuration, with matching OID set to .1.3.6.1.2.1.15.3.1.7 for example. When matching varbinds, NetXMS looks for longest possible match, so it will match OID with any peer IP appended to it, which is fine in that case, because there should be one trap per one peer down event.

Best regards,
Victor

Did you see NetXMS listening on UDP port 162? If not, please post server's log file. If server is listening on port 162, try to set configuration variable LogAllSNMPTraps to 1, and check SNMP trap log. If you will see traps coming, then you just need to configure server to handle traps correctly. You can use this topic as a reference: https://www.netxms.org/forum/configuration/how-to-configure-snmp-traps-from-my-emc-box/.

Best regards,
Victor

You can disable appropriate rule in event processing policy (the one processing SYS_INCORRECT_NETMASK event), or make an exclusion in source field of this rule for problematic nodes.

Best regards,
Victor