NetXMS Support Forum

Hi!

As per subject, I have a problem with the GetSumDCIValue when the result is greater than a 32bit integer.
This is my transformation:

return int64(GetSumDCIValue($node, FindDCIByDescription($node, "WAN: delta in (bytes)"), time()-86400, time()));

It returns 0 when the result cannot be handled by a 32 bits integer.

I have another version of that DCI that uses a different approach to do the same computations and it works properly:

last = int64(GetCustomAttribute($node, "WANCumulativeIn"));
if (last == null)
last = 0;
SetCustomAttribute($node, "WANCumulativeIn", 0);
return last;

The data type assigned to both DCIs is "Unsigned Integer 64 bit" (NB in the last example the custom attribute is updated by another DCI).

Any hints?

Best regards,
Marco

Hi!

Is it supported the "L" extension in the custom schedule (such as in the Quartz java scheduler)? I would like to use a schedule like this:

59 23 L * *

to run a DCI collection at 23:59 in the last day of each month.

Best regards,
Marco

As per subject, it could be useful to have a function to perform the sum of the DCIs for a specified time range (like the GetMin/Max/AvgDCIValue). In my scenario, it will be used to perform cumulative statistics for network traffic (monthly based, with a custom schedule).

Best regards,
Marco

Hi,

on MSSQL (tried on 2012) the GetxxxDCIValue works only with integer values, not with floating points (the wiki states numeric type).

I tried to simulate "manually" the query using this statement (reported in this thread https://www.netxms.org/forum/configuration/getavgdcivalue-not-working/msg11517/#msg11517):

SELECT avg(coalesce(idata_value,0)) FROM idata_<node_id> WHERE item_id=<dci_id> and idata_timestamp between <time_from> and <time_to>

but with floating point values I get the following error message:

Conversion failed when converting the varchar value '14.2999710' to data type int

Is this a restriction that applies only on MSSQL?

Best regards,
Marco

It would be nice to have the possibility to create containers for the dashboards like the container for the nodes or for templates (template group feature) to better organize dashboards when you have a lot of them.

Best regards,
Marco

Is there a way to remove the Max, Avg, and Min values in the line graphs of the dashboards? I like the new info but, in my opinion, there should be the possibility to select if activate or not. If you have a "complex" dashboard with a lot of graphs (like the one attached) the new info will increase the clutter (and doesn't show properly).

Best regards,
Marco

Hi,

I'm trying to use nxapush to push some data to a node. I'm able to use it with a node "directly" connected but I'm in trouble when the node is accessed through a proxy.
Is there any particular instruction to follow when a proxy is involved?

Best regards,
Marco

I grab the Victor's suggestion present in this post:

https://www.netxms.org/forum/general-support/summary-of-network-traffic-1387/msg6048/#msg6048

And I wrote my transformation for the Internal DCI in this way (keeping the original value):

sub getValue(paramDescr)
{
obj = FindDCIByDescription($node, paramDescr);
if (obj != null)
{
   dci = GetDCIObject($node, obj);
   if (dci != null && dci->status == 0 && dci->errorCount == 0)
   {
    val = GetDCIValue($node, dci->id);
    if (val != null)
   return val;
  }
}
   return 0;
}

return $1+getValue("_HF - Download on eth0 (bytes)");


My problem is that I'm getting exactly the value of the original DCI ("_HF - Download on eth0 (bytes)"). Where am I wrong?

Do you have a suggestion on how to reset the counter once per month still mantaining an history (even in another DCI)? The goal is to obtain network traffic statistics per each month.

Best regards,
Marco

Hi,

it seems that this counter:

System.CPU.Count

reports the wrong number of CPU (it reports one less than the real number). I tried it on two Linux distribution (ZeroShell 2.0RC2 and Debian 7 x64). It works properly on Windows (all flavours).

Do it is a bug or I'm doing something wrong?

Best regards,
Marco

I have a node used as a "fake" switch to collect all the traffic generated from a group of virtual machines. That node is configured with the IP 0.0.0.0 so it has no interface. Everything works fine in the data collection, except that the node persists in the "Unknown" status.
Is there a way to force its status to "Normal"? I'm not interested in its "real" state since it hasn't a real state: is just a container for the aggregated values computed from other nodes.

Maybe I'm just using the worst approach to solve this problem, any other suggestion is welcome! :-)

Best regards,
Marco

I did not find a way to re-apply a template to nodes. Is there a way? It could be useful when modifying a template that is assigned to a big amount of nodes.

Best regards,
Marco

I got an error trying to edit a DCI:

An error has occurred. See error log for more details.
For input string: "FF000000"


Any hints for the source of the problem? Where is the error log?

Best regards,
Marco

Hi Victor,

I followed the info here presented:

https://www.netxms.org/forum/configuration/agentless-icmp/msg2273/#msg2273

And I managed it properly when the node is inside the same network where the server is.
What about the possibility to ping a node through a proxy? I tried both method described in the above link but I'm not able to get the correct status, interface is always UNKNOWN and status turns CRITICAL after required number of polls.

Node is properly pingable manually from the node used as proxy (I tried also to add interfaces manually (with MAC address) and I found that they are properly discoverable via the "Find switch port" menu item, but didn't help).

Proxy node works properly, I'm able to query all other SMNP and NetXMS managed nodes.

Any hints?


Here the result of the Poll Node request:

[8/21/12 6:26:16 PM] **** Poll request sent to server ****
[8/21/12 6:26:16 PM] Poll request accepted
[8/21/12 6:26:16 PM] Starting status poll for node TEST-NODE
[8/21/12 6:26:16 PM]       Current interface status is CRITICAL
[8/21/12 6:26:16 PM]       Starting ICMP ping
[8/21/12 6:26:16 PM]    Starting status poll on interface unknown
[8/21/12 6:26:21 PM]       Interface is CRITICAL for 12 polls (10 polls required for status change)
[8/21/12 6:26:21 PM]    Finished status poll on interface unknown
[8/21/12 6:26:21 PM]       Interface status after poll is CRITICAL
[8/21/12 6:26:21 PM] Finished status poll for node TEST-NODE
[8/21/12 6:26:21 PM] Node is still unreachable
[8/21/12 6:26:21 PM] Node status after poll is CRITICAL
[8/21/12 6:26:21 PM] **** Poll completed successfully ****

Best regards,
Marco

Hi Victor,

I have a Windows 2008 R2 cluster with a group of virtual machines managed through Hyper-V hypervisor. Virtual machines are moving between nodes of the cluster depending on load and/or maintenance state of the nodes of the cluster.
I would like to create a pie chart in the dashboard showing the total number of machine in "running", "turned off" and "saved" state for the entire cluster (not for each node). I have exported successfully from each node the relevant values through the WinPerf subagent.
When I tried to add a new parameter to the data collection configuration of the Cluster node, this just "cloned" the parameter to each node, that is not what I need.
Is there a way to get a global value for each data sampled for the entire cluster (summing the values of each node of the cluster)? This will help me to monitor the cluster instead of each node.

Thank you for your help.

Best regards,
Marco

Hi Victor,

I tried to use the web console, I'm unable to go beyond the login page, the service crashes. I tried to execute in debug mode but I didn't get any useful info. Server is Win2003 R2 x64 DB is MS-SQL 2008 R2 SP1 on the same machine.
Do you have any hints?

Best regards,
Marco

Hi Victor,

I have a problem monitoring a DLINK DES-1228P power over ethernet switch. When I enable the SNMP management from NetXMS, after a while (it can be one hour or several hours) some of the ports go down (disconnected) and the physical devices attached to those ports are no more reachable; I have to go to the management interface of the switch and force the ports to go on. This is repeatable, if I put the node in "unmanaged" state everything works fine (all the devices are properly reachable), after switching to "managed" state the same story restart again.
Do you have encountered a similar problem? What can be the reason? Is there any way to instruct NetXMS to reduce polling frequency or wathever else that could help?
Switch is updated with latest firmware and unfortunately is no more supported by DLINK.

Bye,
Marco

Hi Victor,

it could be nice to have the Y-axys value "scaled" to a more readable value like in the legacy console (especially for big numbers). Attachment should better explain.

Bye,
Marco

Hi Victor,

Here some missing features I found using the Java console:

-   It is not possible to select colors for graphs (either background or line color)
-   Preset checkmark for graphs in the right click menu is not shown
-   Preset for graphs is not mantained after saving perspective (windows position is correctly mantained)
-   Predefined graphs prepared with the legacy console are properly drawn but saving the perspective and reloading it at the next run shows only the white background (apparently without any other data inside)

Best regards,
Marco

Hi Victor,

I had a look into the sources for the Android console, I'm not familiar with Android programming but I modified the original version to add a couple of features:

1) When clicking on a notification it goes directly to the alarm view (no more on the main menu)
2) From the alarm list it is possible to jump directly to the last values for the node that has generated the alarm
3) From the last values view it is possible to select one of the last values reported and draw a graph (currently it is possible to select last: 30min, 1hour, 2hours, 4hours, 1day and 1week)

Graph section should be improved but I guess that for now it's enough to show trend of values. It is currently based on the GraphViewLibrary of Jonas Gehring (http://www.jjoe64.com/p/graphview-library.html), published under LGPL.

I don't have much free time but I will be glad to continue if there are no restrictions. Before going on please let me know if my contribution can be useful and how can I send it to you for evaluation.

Best regards,
Marco.

Dear Victor,

I have some problems upgrading from version 1.1.6 to 1.1.8.

1) After upgraded the node I use as proxy, all the nodes accessed through that proxy became unreachable immediately, just reinstalling agent ver. 1.1.6 and everything works properly again. I have the same problem with version 1.1.7. All the nodes accessed directly are fine.
2) After upgraded the server to version 1.1.8, in the legacy console (tab "performance"), for some nodes I'm no more able to see the "Physical Memory" and "Disk Queue" graph properly updated (see attachment). This behaviour isn't related to the version of the agent installed, also this behaviour is present in some nodes accessed directly and in some nodes accessed through the proxy. All my system are Windows based, most of them are x64 but there are also x86 (behaviour present indifferently on both platform).

Here is the config of the agent used as proxy (I removed the public IPs of my server):


MasterServers = xx.xx.xx.xx, yy.yy.yy.yy
LogFile = {syslog}
FileStore = C:\Program Files\NetXMS\var
SubAgent = winperf.nsm
ListenPort = 4700
EnableProxy = yes
EnableSNMPProxy = yes

*WinPerf
Counter = System.IO.DiskQueue.C:"\LogicalDisk(C:)\Current Disk Queue Length":60:A:INT:"Average Current Disk Queue Length on C:"
Counter = System.IO.DiskTime.C:"\LogicalDisk(C:)\% Disk Time":60:A:INT:"Average % Disk Time on C:"
Counter = System.IO.DiskQueue.D:"\LogicalDisk(D:)\Current Disk Queue Length":60:A:INT:"Average Current Disk Queue Length on D:"
Counter = System.IO.DiskTime.D:"\LogicalDisk(D:)\% Disk Time":60:A:INT:"Average % Disk Time on D:"


Do you have any hints for me?

Best regards.
Marco