Comments
-
Thanks, I'll put this on our feature list. I know we looked at doing that but performance was an issue since come configurations can be very large. Rick Matejka
-
Firewalls often have a feature to not decrement TTL and as a result, they will not be seen by NetPath probe. For example, see Firewall/Advanced. Rick Matejka
-
That's good to hear. Thanks for the update. Rick
-
Ahh ok thanks that helps. While there may still be firewall issues, this is an issue with the agent since no path is being returned. Here's a few thing to check: * Enable debugging for NetPath * Check NetPath logs (C:\ProgramData\Solarwinds\Logs\NetPath) for exceptions. If you see "Completed" with Test Results, then the…
-
To verify that the issue is firewall-related, try running tracetcp ( http://SimulatedSimian.github.io/tracetcp) from the probe machine to your target: tracetcp (site):(port) If it fails, then likely your firewall is blocking access to those ports or ICMP replies.netpath What does the graph show exactly?
-
If you select the node, you should see an Issues section in the inspector that will give you the details of the issue. If it's related to packet loss or latency exceeding a threshold, you may want to manually adjust the thresholds to better suite your situation. Instruction can be found in this post: Re: Netpath settings…
-
Johan, Please verify that you are allowing ICMP packets for incoming traffic back to the probe. Rick
-
Not at this time. So how exactly would you want the feature to work?
-
Bill, Those errors are related to the service that looks up BGP data for Netpath. Is it possible you are blocking port 43 to on of the server below? You would need to have access to the first entry and then one of the following entries based on geographic region. Region Hostname Port Allwhois.iana.org43North…
-
You are correct - currently creating new paths is tied to Node Management rights. If you had the option, how would you want to restrict access to the create path function?
-
Greg, That sounds like it could be a bug. If possible, could you export the graph using the directions in this post and then upload the file here? Thanks, Rick
-
What is your Polling Interval?
-
Regarding NetPath, what would be the expected way to get back to the Service Monitors page? Rick Matejka
-
Can you post output from the 'route print' command? Rick Matejka
-
It will depend on how many services you are monitoring, the nature of the service, and your polling frequency. In general the more services you monitor, the more overhead. We monitor usage and based on that, show an estimate to the # of probes you can monitor without a major performance degradation. You will see that when…
-
You can add as many paths as feel your system can support. The system merely provides a guess at how many paths your system can accommodate but it will not prevent you from adding more paths due to resources. Rick
-
Regarding the BGP, see this post: Re: NetPath Errors You need to open access to the server for your locale to NetPath can look up the information about the external nodes.
-
Most likely this is being caused by a proxy. You can try to use tracetcp and see if you get similar results. Rick Matejka
-
On order for NetPath to function, it requires ICMP Type 11 (Time Exceeded) messages to be allowed back to the agent. If those are being blocked, NetPath unable to locate the intermediate nodes. Can you see what TraceTCP shows? Rick
-
So by "no information", do you mean that when you view the path you only see the error messages stating that no data is available? Can you verify that the Netpath plugin was correctly installed? Instructions can be found in this post: Netpath Rick
-
Unfortunately not at the moment. What information and in what format would you find most useful? Rick Matejka
-
This is a known issue an we are working on a fix. Rick Matejka
-
Hping and traceroute will show results for a single path. NetPath monitors the service from multiple paths and then aggregates the results. It's possible one of the paths has higher latency than the others. If you can share a screen cap of the graph, that would help to see what could be going on. Rick Matejka