NetXMS Support Forum

if you are using deployment tools then you should already have everything you need to deploy.  It shouldn't take to much to copy a folder to the machine and create a shortcute to nxmc.exe to the users desktop. I have used PDQDeploy to do the exact same thing without issues

Victor,

I have been going back and forth with them trying to have them fix the inconsistencies to no avail see their response below.

We have looked into the SNMP frames to find which OID NetXMS is polling from EDR-810 during auto-topology. We found out that modifying the subtype in LLDP does not fix the topology issue of NetXMS in our test. However, after discussing with PM, making EDR-810 compatible of NetXMS would be not in the specs and would be a CV request. In addition, we found that NetXMS polls dot1dBasePortIfIndex from EDR-810. However, the EDR-810 does not support dot1dBase-MIB. We have confirmed the auto topology can be done for EDR-810 with FW v5.6 in MXView

Is the fact that the switch does not support dot1dbaseportindex really matter.  Looking at the trial version of MXView  it doesnt seem to do anything more than what Netxms does and pretty much queries the same mibs except dot1dbase.  But it is able to find the topology no problem.   If this is somthing im just going to have to deal with. Is it possible to provide me with an example of a HOOK i can use to try and change the names of the interfaces after a poll. I have renamed the interfaces manually and the topology does seem to be found but it is very inconsistant every other poll will not find connections and it will never find a netxms agent that is connected to any port on the switch. However it does list the connections via info -> lldp but the port number is always off by one.  For example the agent will be connected in port 4 but lldp lists it as port 3.  So i do think there is some netxms lldp related issues going on.


Thanks

Victor,

Is there anything I can do to provide a workaround to this on the netxms side of things? can I force the naming of these interfaces to match what is being sent by lldp.  I was given a newer firmware that was supposed to fix a known issues but I am still not seeing the topology in Netxms .  However The trial version of the vendors NMS works just fine.  But would rather not pay for an inferior product that doesnt have half the features of Netxms.

Victor,

I attached the ifxtables in the previous message. Another thing I noticed is the FDB table does not list an interface. Is this also a result of the behavior of the switch. 

"MAC Address","Port","Interface","VLAN","Node","Type"
"CC:4E:24:BC:D1:2C","9","

• ","1","TMC-ICX-CORE","dynamic"
  "AC:CC:8E:AC:01:80","2","
• ","1","172.27.x.x","dynamic"
  "00:40:8C:E6:D6:37","9","
• ","1","172.27.x.x","dynamic"
  "00:04:81:02:1E:EB","10","
• ","1","","dynamic"
  "00:04:81:02:1F:9B","10","
• ","1","","dynamic"
  "00:04:81:04:1B:DB","1","
• ","1","","dynamic"
  "00:04:81:04:1D:AF","10","
• ","1","","dynamic"
  "00:04:81:04:4F:19","10","
• ","1","","dynamic"
  "00:14:7B:20:20:8C","9","
• ","1","","dynamic"

Here you go. At this point I dont think this has anything to do with Netxms but any help would be much appreciated.  I have also emailed the switch support since the switch is clearly  advertising values that are not even configured in the tables.  I have a feeling this maybe on purpose since this particular vendor has a paid version of their own NMS
.


Thanks,

Gary 

I have attached the logs.  I think i know whats going on and it seems to be on the switch side but not 100% sure.  For each node that has been fully polled the interfaces all show as

ETHERNET PORT 1
.
.
ETHERNET PORT 10

However the lldptables from the switch show completely different port names

PORT1
.
.
PORT8
SFP-1GLXLC-T
SFP-1GLXLC-T

I can only get the topology to work if I uses alias's and SNMP SET the alias to match the lldp port name.  So setting port 9 and 10 to SFP-1GLXLC-T  on both switches will allow the connections to be found.  But I really don't want to script this up just to get things working and  if the SFP is changed it could end up being a different port name.  I would think Netxms would be able to just use the port index and FDB but maybe since LLDP is enabled that is not being used.   One other thing is the switched is a L3/L2 but is only on layer2 mode but is still discovered by netxms as being a router so the dot1qtpfdbtable does not return any entries only dot1dfdtable returns. Hope this helps and thank you for your time.

Gary 

At work we switched to a different set of l2 switches and I do not get any topology info at all. But if I right click ->  info -> lldp I can see the neighbors. I can alos mib browse and see both local and remote lldp tables populated.  I though this all netxms needed to build the topo. The brand of switch is Moxa EDR 810 if that helps

I noticed the in 3.2.350 that there is a tsdb.sql init file  but when building from source tsdb is not an optional driver to install.  What is the standard install for timescaledb.   I am assuming i would follow the timescale.com docs to install postgres and the timescaledb plugin but for netxms would I then only need to init the tsdb.sql file or does both pgsql and tsdb neet to be initialized.

Thanks

Working now ... Thank you

I tried directly on the server which has version Java Version 1.8.0   and still nothing.  However this scripts works just fine on both server and separate client.

import sys
from fnmatch import fnmatch

if len(sys.argv) > 1:
  mask = sys.argv[1]
else:
  mask = ".1.3.6.1.*"

nodes = filter(lambda x: isinstance(x, objects.Node), s.getAllObjects())
matched = filter(lambda x: fnmatch(x.getSnmpOID(), mask), nodes)
for node in matched:
  print node.getObjectId(), node.getObjectName(), node.getPrimaryIP().getHostAddress()



any script with getAllChilds method does not work 

something has definitely changed the older api shows getChilds() method the latests does not  and I cant even find the documentation for my version 2.2.16

https://www.netxms.org/documentation/javadoc/versioned/Release-2.2.8/

Where there any changes to the nxshell ?  Im using 2.2.16  and all my old scripts no longer work. I even tried most of the examples in the wiki and they no longer work either.  This is the error i receive

WARNING: An illegal reflective access operation has occurred
WARNING: Illegal reflective access by org.python.core.PySystemState (file:/home/gmonk/Development/Python/Netxms%20Scripts/nxshell-2.2.16.jar) to method java.io.Console.encoding()
WARNING: Please consider reporting this to the maintainers of org.python.core.PySystemState
WARNING: Use --illegal-access=warn to enable warnings of further illegal reflective access operations
WARNING: All illegal access operations will be denied in a future release
node_id,interface_id,name,ip,mask
Traceback (most recent call last):
  File "iftest.py", line 8, in <module>
    allInterfaces = node.getAllChilds(objects.GenericObject.OBJECT_INTERFACE)
AttributeError: 'org.netxms.client.objects.Node' object has no attribute 'getAllChilds'

Great    Thanks for the info

If I want to migrate from Mysql to Timescale is there any way to import the data from mysql ?  By the way awesome choice   I always thought timeseries databases where better suited for NMS's   


Thanks

.1.0.8802.1.1.2.1.4.1.1.2.84659416.13.1 [INTEGER] = 13
.1.0.8802.1.1.2.1.4.1.1.2.84658408.12.2 [INTEGER] = 12
.1.0.8802.1.1.2.1.4.1.1.3.84659416.13.1 [INTEGER] = 1
.1.0.8802.1.1.2.1.4.1.1.3.84658408.12.2 [INTEGER] = 2
.1.0.8802.1.1.2.1.4.1.1.4.84659416.13.1 [INTEGER] = 4
.1.0.8802.1.1.2.1.4.1.1.4.84658408.12.2 [INTEGER] = 4
.1.0.8802.1.1.2.1.4.1.1.5.84659416.13.1 [STRING] = ec-13-b2-63-96-7c
.1.0.8802.1.1.2.1.4.1.1.5.84658408.12.2 [STRING] = 6c-3b-6b-54-b0-e2
.1.0.8802.1.1.2.1.4.1.1.6.84659416.13.1 [INTEGER] = 5
.1.0.8802.1.1.2.1.4.1.1.6.84658408.12.2 [INTEGER] = 5
.1.0.8802.1.1.2.1.4.1.1.7.84659416.13.1 [STRING] = Port 14
.1.0.8802.1.1.2.1.4.1.1.7.84658408.12.2 [STRING] = ether1
.1.0.8802.1.1.2.1.4.1.1.8.84659416.13.1 [STRING] = Port 14
.1.0.8802.1.1.2.1.4.1.1.8.84658408.12.2 [STRING] =
.1.0.8802.1.1.2.1.4.1.1.9.84659416.13.1 [STRING] = 1971_BPDO_SW
.1.0.8802.1.1.2.1.4.1.1.9.84658408.12.2 [STRING] = MTIK_UNDO_AP
.1.0.8802.1.1.2.1.4.1.1.10.84659416.13.1 [STRING] = Netonix WS-12-250-AC
.1.0.8802.1.1.2.1.4.1.1.10.84658408.12.2 [STRING] = MikroTik RouterOS 6.42.9 (long-term) RB Groove A-52HPn r2

The above is the lldpRemTable from a Netonix switch in the field.  We are having the same issues with our Mikrotik switches.  I understand the need for the LLDP chassis ID but it seems there are several switches that only populate remote table info  and are there any cases why the LLDP chassis ID would differ from the Bridge Base Address could you not supplement this base address assuming the checks for LLDP chassis ID fail but only if there are entries in the LLDPRemtable

Thanks,

Gary

Victor,

Sorry i completely forgot about this.  No this switch does not respond to those OID's only the remote table seems to be populated. Is it possible to do a check on the remote table for any entries and enable LLDP  based on if any neighbors are found as well.   I dont know why this switch does not respond to .1.0.8802.1.1.2.1.3.2.0  or  .1.0.8802.1.1.2.1.3  but Im sure there could be other off brand switches that do the same.


Thanks,

Gary