Netonix Forums

Eric Stern wrote:You'll need to figure out why the connection to your switches is timing out. No one else has reported this problem, it has to be something specific to your network.

Nope. It happens to me CONSTANTLY, as well. I finally disabled the software. My system is not going down. I stream radio with 0 buffer and the audio hasn't so much as glitched, but the software has sent out about 20 emails in the last 24 hours that "x radio is offline, x radio is online"

Nope. It happens to me CONSTANTLY, as well. I finally disabled the software. My system is not going down. I stream radio with 0 buffer and the audio hasn't so much as glitched, but the software has sent out about 20 emails in the last 24 hours that "x radio is offline, x radio is online"

thanks for your post, we will look into this...

Am also seeing lots of ETIMEDOUT and ESOCKETTIMEDOUT emails. It mostly happens on switches that have a few hops to get to from the server, but there are a couple of times that the switch at the other end of our rock solid one-hop 600mbps link has timed out too

Maybe there could be a (configurable?) number of failures before the devices is deemed 'offline'.

scottb wrote:Am also seeing lots of ETIMEDOUT and ESOCKETTIMEDOUT emails. It mostly happens on switches that have a few hops to get to from the server, but there are a couple of times that the switch at the other end of our rock solid one-hop 600mbps link has timed out too

Maybe there could be a (configurable?) number of failures before the devices is deemed 'offline'.

What version of firmware are you running on the switch and what version of Manager?

I use the Manager at my WISP and I have a router and switch at every hop and some switches are 6 hops out and I do not see this behavior.

firmware 1.4.7 and manager 1.0.7.

474 offline/online notifications from 19 monitored switches since Monday :)

it's obviously just some kind of packet loss and I'd attribute it to the wireless gear rather than the switches. It's not affecting our regular data and I'm not really worried about it, but making manager try a couple of times before marking the switch offline would sort it out.

I love the idea of this program. However I cannot get an address range to discover and add any switches. If I type the whole IP it will add the switch. But /12 and /24 did not find a single switch to add. I've tried 1.0.7 on debian and windows server. Tried adding multiple ways 172.16.0.0/12, 172.16.15.0/24, 172.16.15.0/255.255.255.0

I've had another pass at the Netonix Manager and have a few (minor) suggestions.

I had previously experimented with version 1.0.5 (in July). On 11/1/2017, I installed latest Node.js, then installed Version 1.0.7 on a Windows 10 box and started it up at ~3:45pm. Here are some things I've noted.

1. On startup there were a set of log messages from September (see image 1 below). I'm not aware of having done anything with this software in September.

2. A minor annoyance: The scanning message shouldn't overlay the log scrolling area! (also visible in Image 1 below).

3. I started a scan of 10.8.0.0/16. It found three devices rather quickly but clearly had a long way to go. On the assumption this would take many hours but would finish before I returned, I added another scan (for 10.12.0.0/18). This caused the # of devices found to reset to zero even though the original scan was obviously proceeding and the devices display still shows the three devices found so far. (Images 2 & 3 below). When I came back, the 10.8.0.0/14 scan had found all 51 switches that are actually in the range 10.8.4.5 thru 10.10.211.2.In addition, the scan of 10.12.0.0/18 had also found 5 switches in the range 10.12.25.3 thru 10.12.35.2. So I now have 56 switches in the "Devices" list but the log is as shown in image 4 below. I conclude that you can add an additional scan before an earlier one has ​completed and it will work but the count will be short by however many devices had already been found when the second scan is initiated, i.e. defining a new scan resets the counter even if an existing scan is in progress.

4. A day later I've scanned my whole network. I should have 156 Netonix devices but there is no easy way to get a count of what the manager has found, short of scrolling and counting lines (some single, some double) on the screen.

5. Now having counted, I see the Netonix Manager has only found 134 devices. I don't see a way to print out a list as my number of devices far exceeds the window in the web page, i.e. when I print the web page I only see the dozen or so radios visible at any one time, not the whole list. So now I will have to scroll through comparing entries to my NetXMS list (which I can print). Even better than printing might be the ability to export the Device list to a .csv file. Then I could print it and, even better, format it for convenient DIFs against the NetXMS output.

6. I like the Manager's ability to display log alerts but the display window that comes up is too narrow and can't be widened (see image 5) enough to display the whole alert.

In any event, thanks for the work so far. This looks useful in general and certainly looks like the solution the next time I have to upgrade firmware.

Thanks,
Brough

One of our guys killed our Internet access briefly so I lost my VPN tunnel into our network and everything in the Netonix Manager turned red. I re-established the tunnel about a minute later and things started turning green, however, some devices appear to have lost the device names during this interruption. It's been five minutes since connectivity was restored. Everything turned green (or yellow) very quickly but these five devices are still without names:



Any ideas?
How do I get the names back?

Now I've started looking at which devices the Netonix Manager missed in it's scan. I can see no pattern. There were two switches at 10.28.48.2 and 10.28.48.3 which should have been found by the scan of 10.28.0.0/15. Other devices on either side were found.

And when I did a scan of 10.28.24.0/24, it immediately added these two devices to the device list.

The Windows 10 box that is running the Netonix Manager is connected to our network by a VPN tunnel, but this tunnel does not auto-reconnect. If it had dropped, I would have had to manually restore it.

So now I just have to continue comparing the screen display to my NetXMS printout and re-scan those parts of the address space that have been skipped. Presumably I will eventually get from 134 devices found in the first passes to the actual number which is 156 devices.

On the 5 devices which have lost their names... I tried re-scaning the address range of one of the devices. That did nothing. I then deleted one of the devices that had no name and then re-scanned the /24 address block it was in. It was found immediately and has been restored. Of course I can do this for the other four devices as well, but it would be nice to know how this happened and how to avoid it in the future.