Greetings,

Supremely awesome program. I am sending a trace every 30 seconds to mainly track pl. I would like to email the plot results on a regular basis like every 4 hours. What I am having difficulty understanding is how to setup the conditions. Conversly, I would like to also send an alert as soon as a pl occurs, which the program does track each pl error count which is neat in itself.

Is the entire plot sent in the email? or only the traces to examine. Like, if its examining the last 10 traces, if a condition is met, only those 10 traces are sent or the entire plot? Looks like the entire plot is. The second problem is I ran a test and the attachment, as well as autosave results, does not appear to include packet loss data in the graph, only in the text portion of the email. Would be nice if the saved data file contains the pl. Also, the text report in the email is exactly what I need, but cant seem to duplicate it in the main program as well, the export process is not the same, nor is copy as summary text.

As a test, this is how I have mine setup. Traces to examine = 1000, alert when 1 trace exceeds 1ms. Send email alert each time condition is met, max wait = 5 minutes. I was trying to get it to email me every 5 minutes with the full plot results. I am not receiving emails. Test works fine.

Thankyou again for a great program.
Jason

Hi, Jason.

The data in email is a data file that includes all data, including packet loss and good data. It exactly represents recent data in memory - the *quantity* of data represented is based on when the alert conditions started, and when the last email went out.

What you're asking for should be possible. You can set up two separate alerts:

One alert always fires (ie: when 1 or more of the last 5 samples is over 1ms), and set it up to email only once every 4 hours. You'll get 4 hours of data in each email (data up to the previous alert email that was sent, plus just a bit of overlap). The only goal of this alert is to email you data occasionally.

The second alert you set up for packet loss or latency (whatever conditions you need), and have it email immediately. This alert will include samples only for the period when alert conditions were happening, plus just a bit of time prior to that.

Then, tie both of these alerts to the IP Address you want to monitor (likely your final destination in this case) by tracing to that address, right-clicking the hop you want to "Watch", and then selecting "Watch this host (Alerts)..." from the menu. This will set up the alerts to watch that host.

The alerts make their own decisions about what data to include based on the conditions you configure (including the Traces to Examine), the last email that went out, and whether the alert is configured to send an email immediately or wait for a while.

Your alert configuration sounds good - you probably just missed the part where you need to tie the alert to a host (see two paragraphs up here).

The auto-save results are done completely separately from the alerts. Alerts don't check this or use this configuration in any way. You should be able to get the export (File / Export to Text File) to output data very similar to the alert email's data - where an * is a lost packet. If you can't, please send an email to support@pingplotter.com with examples of both and we can help get things working for you.

- Pete

Great, thankyou for your reply.

I didnt catch the step about assigning an alert to an address. Doing so, alerts are working fine now. What I would like to do, is if any hop anywhere in my trace, if anyone has a problem, it will run an alert. I have noticed that certain hops will timeout and have pl, while the destination may not. This is oposed to setting an alert for each hop, which, is not that big of a deal anyway, which is what I will be doing.

Not sure what was wrong originally, but on my recent emails, I do see pl counters. Maby when selecting the test button, it has this issue, working fine now.

Again, this is a great program and very usefull.
Thankyou
Jason

Hi, Jason.

You don't want to monitor intermediate hops, to be honest. The latency or packet loss displayed in the intermediate hops is unimportant unless it translates to the same at the final destination.

Remember that every packet needs to go through each hop on the way there - and if each router is passing data on to the next router quickly and efficiently, the latency / packet loss of packets that have TTL=0 *at that router* isn't important.

The data for intermediate hops is only interesting in how it affects the final destination. If you're getting packet loss and latency at the final destination, then you can look at the intermediate hops to see which router is introducing that. If, however, the final destination is performing well, you shouldn't worry much about the intermediate hops.