NetXMS Support Forum

Hi,

To manually add nodes, right click on the "All Services" object in the Object Browser, which should be on the left-hand side of the window (If not, press F3, or look under the View menu to open it). Then select "Create node..." from the dropdown menu. You can also create Containers this way to organize your nodes.

The node probably appears automatically after a while because network discovery is enabled. You can find its options under the Configuration menu.

The user manual is unfortunately a bit outdated, the wiki contains some more up to date information but isn't quite complete yet either: http://wiki.netxms.org

Good luck!

Sounds good!

I look forward to the next release(s)!

If it were implemented like the security tabs on objects you could use groups for it, yes.

Keep in mind events and alarms are two seperate things though; events come from agents and DCI monitoring and don't really have any user interaction, alarms come from the event processing policy.
Although each alarm has an event at its root, a single event might have to create different alarms for different target groups (like when the server goes down, both SA's and DBA's have to get an alarm). Security settings at the event template level wouldn't be very suitable for this, I think.

I did. What I want to achieve is that, on for example a SQL server, SQL admins can see alarms about the SQL server (Page life expectancy, etc) and system admins can monitor system related alarms (temperature, C drive free space, ...), without having the other one's alarms clouding the console.

As I said, different people, but on the same nodes.

Crashed again saturday afternoon. Last messages in log file were
[06-Oct-2012 16:08:27] Thread "Syncer Thread" does not respond to watchdog thread
[06-Oct-2012 16:08:27] Thread "Poll Manager" does not respond to watchdog thread

[06-Oct-2012 16:09:36] Log file opened
[06-Oct-2012 16:09:37] Database driver "mssql.ddr" loaded and initialized successfully
[06-Oct-2012 16:09:39] Stalled database lock removed
[06-Oct-2012 16:09:40] Unable to load module "": The specified module could not be found.
[06-Oct-2012 16:09:40] Network device driver "BAYSTACK" loaded successfully
[06-Oct-2012 16:09:40] Network device driver "CATALYST-2900XL" loaded successfully
[06-Oct-2012 16:09:40] Network device driver "CATALYST-GENERIC" loaded successfully
[06-Oct-2012 16:09:40] Network device driver "CISCO-ESW" loaded successfully
[06-Oct-2012 16:09:40] Network device driver "DELL-PWC" loaded successfully
[06-Oct-2012 16:09:40] Network device driver "ERS8000" loaded successfully
[06-Oct-2012 16:09:40] Network device driver "NETSCREEN" loaded successfully
[06-Oct-2012 16:09:40] Network device driver "PROCURVE" loaded successfully
[06-Oct-2012 16:12:12] Unable to bind socket to port 58942 in function LocalAdminListener: Only one usage of each socket address (protocol/network address/port) is normally permitted.
[06-Oct-2012 16:12:12] Unable to bind socket to port 162 in function SNMPTrapReceiver: Only one usage of each socket address (protocol/network address/port) is normally permitted.
[06-Oct-2012 16:12:12] NetXMS Server started
[06-Oct-2012 16:12:12] Unable to bind socket to port 4701 in function ClientListener: Only one usage of each socket address (protocol/network address/port) is normally permitted.
[06-Oct-2012 16:12:12] Unable to create socket in function ClientListener
[06-Oct-2012 16:12:52] Thread "Item Poller" does not respond to watchdog thread

I had to reboot the server to free up the ports, server is running fine again since. I also enabled the CreateCrashDumps option in netxmsd.conf, if that still exists, and will send you the crash dump if it occurs again.

Hello Dave,

I've tried out
if ($1 > 5000)
  return 0;

on a test DCI (on v1.2.3, though), and it seems to work as expected.
Are you sure the script is applied to radiusAccServTotalResponses, and not just only to the Requests DCI?
Also keep in mind that the script does not work retroactively; it will only change the results of the DCI after the change was made.

If you can post some screenshots of the General and Transformation tabs of the DCI, maybe we can find something out of the ordinary there.

Hi,

Would it be possible to define a specific group of users that can view an alarm in the NetXMS console/web interface?
This would allow a single NetXMS server to be more easily used by multiple people, but with different interests on the same nodes.

Perhaps even some sort of security settings for alarms, so you can select who can view/acknowledge/resolve/terminate alarms coming from a specific event processing rule? These permissions could then be ANDed with the existing permission set for alarms at node level.

Hi,

We haven't had a crash since my last post, so it appears that v1.2.3 is running more stable.

I will keep you informed if it does happen again.

Hi,

We've been having some problems with our NetXMS environment since going from ~200 nodes to above 400 on v1.2.0, and currentely 787 on v1.2.3. The server crashes at random times, every 2-10 days. Crashes have occured with and without users logged on.

NetXMS runs on a 32bit Windows 2003 server with a MS SQL Express 2008 R2 database on the same server.

At first, we would see network service objects changing status to "unknown" up to an hour before the actual crash. The Windows event log shows the agent reporting
"Communication session broken: An existing connection was forcibly closed by the remote host."
Until finally the server crashed and had to be restarted. The server log shows no errors.

More recently, the v1.2.0 server crashed again but this time without the agent errors. The event log shows there were no SYS_SERVICE_UNKNOWN events prior to this crash either.

I hoped an upgrade to v1.2.3 would resolve this problem, however the server crashed again tonight with no users logged on, and no logged SYS_SERVICE_UNKNOWN events, after having upgraded Monday.

I'll send the Dr. Watson crash dump to [email protected]. If I can provide any other information, feel free to ask.

Hi,

A strange thing we've been having for quite a while is that mails generated by NetXMS show the wrong timestamp:
For example, an event that occurs at 8:46, will show as if it was sent at 9:46 (but will arrive at 8:46).

I'm in CET (UTC+1), but during the summer, we have CEST. (UTC+2) I did some research as to what was causing this strange issue, and found the following in the header of the e-mail:
Received: ... Wed, 1 Aug 2012 08:46:02 +0200
Date: Wed, 1 Aug 2012 08:46:02 +0100
We have v1.2.0 running on a 32bit Windows 2003 server, time zone is set to GMT+1 with DST enabled.

I traced this back to server/core/email.cpp, line 273 to 277. The neccessary time zone information is collected via GetTimeZoneInformation, but daylight savings does not seem to be taken into account. As far as I understand, the return value of GetTimeZoneInformation can be used to determine if standardBias or DaylightBias should be taken into account (http://msdn.microsoft.com/en-us/library/windows/desktop/ms724421%28v=vs.85%29.aspx )

Hi,

I think you're looking for the event/snmp trap/syslog logs, rather than the monitors.

Look under the View menu in the Java console.

You may have to define a time frame for which to view the log for it to be useful, though.
Click "Add column" under Condition, and select "Time" to do so.
After filling in the desired time frame, click the "Execute" button (green/white "play" arrow) in the top right corner.

Hi,

Could there be timing issues with this approach? If you have 10 nodes dependant on 1 switch, all with a status interval of 1 minute, you're bound to have some nodes going down before the switch node.

If it's turned in to a full-fledged rule, it'd be really cool if the rule could trigger the status (or more generally, any random DCI?) to be checked before evaluation; but perhaps to prevent it from bogging down the event processor make this optional.

Hi,

I've had this same error in our W2K8 Citrix environment.
It appears to be caused by the working directory path in the nxmc.ini file:

-data
@user.home/.nxmc/data

@user.home should refer to the current user's profile directory, but this appears to be broken in some cases. After replacing this with a different path the issue was solved for us. I believe it's related to
http://bugs.sun.com/view_bug.do?bug_id=4787931

We have the my documents folder of each user mapped as a network drive, so we changed it to place the .nxmc folder onto this drive.

Thanks for the excellent information!

Is there any way to set the expected state for a selection of interfaces? Opening them one by one and adjusting it can get a bit tedious.