NetXMS Support Forum

Thanks for the reply. There are two issues I have:

• First the Synology device is not on the same network as the netxms server. I am using another workstation on the network of the Synology NAS as an SNMP proxy to monitor the Synology NAS.
• I can not install the netxms agent on the Synology NAS because it is running a custom linux distro that is not supported by netxms agent (I tried)

And so I need some way to TCP proxy from the workstation that is on the same network as the Synology NAS to the Synology's web admin panel on port 5000.

I can enable TCP proxy on the workstation agent however there seems to be no way in object tools to configure the tool to access a specific IP along with the port.

I have used this feature in other remote access software which would allow specifying both the Remote Host IP and Remote Host Port to proxy. Not sure if this is possible with the TCP proxy feature or if there are any workarounds.

So I've been playing around with the new TCP Proxy feature and it seems I am only able to access local ports on a particular node. Is it possible to proxy to a service located on another device on the same network as the node?

For example I would like to access a Synology web admin panel through another node on the same network. The local network address for the admin panel would be something like http://192.168.1.10:5000 . When configuring the Local Command in Object Tools the TCP Tunnel check box only allows to specify a port number. There seems to be no where to supply an address. Is it possible to configure the current TCP proxy to do this or is it only possible to proxy local ports on a node?

I've used remote access software in the past that allowed me to do this. Not sure if I'm missing something here.

Quote from: djex81 on April 13, 2015, 06:57:32 PM

Quote from: djex81 on April 11, 2015, 02:20:22 AM
Hi Victor, in this new update you fixed the broken SNMP in agents and that works great but now I am unable to use SNMP from the server netxms is installed on. It was working in version 2.0-M2. What happens is this:

1. I do a configuration poll on the netxms server node. SNMP is connected to and values read.
2. I try to do a SNMP walk from netxms on the local netxms server node and it fails or gives no results back.
3. If I then do another configuration poll on the netxms server node SNMP fails and isn't found.

I know SNMP works as I can SNMP walk with a separate program without any issues. I tried deleting the netxms server node and having it recreate it self but that did not help. Could it be something to do with the new IPv6 interfaces in this new version?

Basically, SNMP is working for all other nodes except the node netxms server is installed on.

To add to my previous post quoted above when I do a status poll on the netxms server node SNMP errors with:

[13.04.2015 11:50:24] Checking SNMP agent connectivity
[13.04.2015 11:50:32] SNMP agent unreachable

The SNMP agent is local so I'm not sure why it's unreachable. I can sometimes get it to connect to SNMP but most of the time it will not. I find if I am able to connect and I try to do a SNMP walk from netxms it will loose SNMP connection and return no results. I have had it return results successfully but only a few times. What debug level should I use so I can check the logs? I tried debug 7 but I couldn't see much in the way of SNMP errors.

UPDATE: I believe I've found the cause of the problem. It turned out to be the agent for our APC UPS causing SNMP to hang randomly. I updated the APC agent and so far it's working well. It's strange that it was working just fine in 2.0-M2 though. Oh well it's working now.

Quote from: Victor Kirhenshtein on April 14, 2015, 09:39:27 AM

What operating system and SNMP version used? If you are using SNMP v3, do you use authentication and encryption, and what algorithms?

Best regards,
Victor

I am using Windows Server 2008 R2 Standard with SNMP version 2c. There is no authentication or encryption at the moment.

Attached are screen shots of the properties pages for the server node.

Quote from: Victor Kirhenshtein on April 13, 2015, 09:14:15 PM
What address you have set as primary host name for management server node? Can you do nxsnmpwall on that address?

Best regards,
Victor

For the server node I am using the local address 10.0.0.2 . Yes, I can nxsnmpwalk on 10.0.0.2 without any issues.

Quote from: djex81 on April 11, 2015, 02:20:22 AM
Hi Victor, in this new update you fixed the broken SNMP in agents and that works great but now I am unable to use SNMP from the server netxms is installed on. It was working in version 2.0-M2. What happens is this:

1. I do a configuration poll on the netxms server node. SNMP is connected to and values read.
2. I try to do a SNMP walk from netxms on the local netxms server node and it fails or gives no results back.
3. If I then do another configuration poll on the netxms server node SNMP fails and isn't found.

I know SNMP works as I can SNMP walk with a separate program without any issues. I tried deleting the netxms server node and having it recreate it self but that did not help. Could it be something to do with the new IPv6 interfaces in this new version?

Basically, SNMP is working for all other nodes except the node netxms server is installed on.

To add to my previous post quoted above when I do a status poll on the netxms server node SNMP errors with:

[13.04.2015 11:50:24] Checking SNMP agent connectivity
[13.04.2015 11:50:32] SNMP agent unreachable

The SNMP agent is local so I'm not sure why it's unreachable. I can sometimes get it to connect to SNMP but most of the time it will not. I find if I am able to connect and I try to do a SNMP walk from netxms it will loose SNMP connection and return no results. I have had it return results successfully but only a few times. What debug level should I use so I can check the logs? I tried debug 7 but I couldn't see much in the way of SNMP errors.

Hi Victor, in this new update you fixed the broken SNMP in agents and that works great but now I am unable to use SNMP from the server netxms is installed on. It was working in version 2.0-M2. What happens is this:

1. I do a configuration poll on the netxms server node. SNMP is connected to and values read.
2. I try to do a SNMP walk from netxms on the local netxms server node and it fails or gives no results back.
3. If I then do another configuration poll on the netxms server node SNMP fails and isn't found.

I know SNMP works as I can SNMP walk with a separate program without any issues. I tried deleting the netxms server node and having it recreate it self but that did not help. Could it be something to do with the new IPv6 interfaces in this new version?

Basically, SNMP is working for all other nodes except the node netxms server is installed on.

Ok so I've come up with a temporary fix for the XMPP delay issue. I found a python library for sending messages via XMPP and it works quite good. I tested it on Windows but it should work on linux and unix. Below are the instructions for Windows.

1. Download and install Python 2.6.6 from here: https://www.python.org/download/releases/2.6.6/
2. Download and install xmpppy 0.4.0 from here: http://downloads.sourceforge.net/project/xmpppy/xmpppy/0.4.0/xmpppy-0.4.0.win32.exe
3. Create a folder any where to hold the sendXMPP.py script. I used C:\NetXMS\XMPP_FIX\

File: sendXMPP.py (not my code, its modified example code on the xmpppy project site)

#!/usr/bin/python
# \$Id: xsend.py,v 1.8 2006/10/06 12:30:42 normanr Exp \$
import sys,os,xmpp,time

if len(sys.argv) < 2:
print("Syntax: xsend JID text")
sys.exit(0)

tojid=sys.argv[1]
text=' '.join(sys.argv[2:])

jidparams={}

jidparams['jid']='[email protected]'
jidparams['password'] = 'LOGN_PASSWORD'

jid=xmpp.protocol.JID(jidparams['jid'])
cl=xmpp.Client(jid.getDomain(),debug=True)

con=cl.connect(('YOUR_XMPP_SERVER_HERE',5222),  use_srv=False)

if not con:
print("could not connect!")
sys.exit()
print("connected with",con)
auth=cl.auth(jid.getNode(),jidparams['password'],resource=jid.getResource())
if not auth:
print("could not authenticate!")
sys.exit()
print("authenticated using",auth)

# cl.SendInitPresence(requestRoster=0)   # you may need to uncomment this for old server
id=cl.send(xmpp.protocol.Message(tojid,text))
print("sent message with id",id)

time.sleep(1)   # some older servers will not send the message if you disconnect immediately after sending

cl.disconnect()


4. Take the above code and create a sendXMPP.py file in the folder created in step 3. Make sure to edit the script to use your own login, password and xmpp server.
5. Inside netxms add an action and fill out the fields as follows:

Name: External Send XMPP
Type: Execute command on management server
Command: C:\\Python26\\python.exe C:\\NetXMS\\XMPP_FIX\\sendXMPP.py "ACCOUNT_TO_SEND_TO" "%m"

Note: Your python path and sendXMPP.py path might be different so change the paths in the command accordingly.

6. Now add the action you just created in the desired event processing policy and it should be ready to go.

If you want to send to multiple accounts at once you will have to either create multiple actions for each account or modify the python script.

I don't have time at the moment to try it on linux. Maybe you could give it a go ofca.

Quote from: ofca on April 10, 2015, 03:49:46 AM
I'm using ejabberd, not that it matters. Since everybody using XMPP seeem to have this problem, I strongly believe that main problem with XMPP support in NetXMS is Victor not using XMPP

Yea I'm using talk.google.com for now. I used XMPP in place of SMS notifications on my mobile phone. Works well besides the 5 - 7 min delay of the event processor in netxms. Maybe I'll work on a quick external program to send the XMPP alerts until this is fixed. Are you using Windows or Linux server for your netxms server?

Hi Guillaume,

Did you read this forum post? https://www.netxms.org/forum/configuration/web-site-monitoring/

Try this. Set the Response to use ^HTTP/1\.[01] 200.*

There might be an issue with it finding the OK on the testDB.php response.

Has any progress been made on this issue?

I am having this exact issue as well. ~5 min delay with the event processor stuck at 1 until the message is received in the XMPP client then it clears to 0. Sending commands to netxms via XMPP works instantly.

Windows Server 2008 R2 Standard.
NetXMS version 2.0-M3

What XMPP server is everyone using?

Thanks for the info. I removed some MIBs that I did not need and used the -z option and now the netmxs.mib files is ~5mb where it use to be ~45mb before. Works good now.

I just updated to version 2.0-M3. I noticed "Loading MIB file on startup" does not complete and slows down the system. I did have custom MIB's installed so I removed them and restored the MIB's folder back to the originals and recompiled but that still didn't fix the issue.

UPDATE: This issue only seems to happen with a remote instance of the management console. The local console on the server works fine.

Quote from: Victor Kirhenshtein on April 07, 2015, 08:59:49 PM
Hi,

SNMP proxy functionality was broken in 2.0-M2. I've fixed it in 2.0-M3. I've also change logic a bit so if node is proxy for itself it will use loopback address instead of primary IP address for SNMP requests - this can help avoid possible firewall problems and secure SNMP access to node by allowing only requests via loopback.

Best regards,
Victor

Hi Victor, Thank you for fixing this issue. SNMP in 2.0-M3 works great.

Does any one know if this is possible?

I found if I add a second node with the IP 127.0.0.1 and setting Host B as the SNMP proxy I can access Host B's SNMP. But this isn't ideal as I now have two nodes for Host B. One for the NetXMS agent and one for SNMP.

I am using version 2.0-M2. Is this issue possibly a bug?