[smokeping-users] best test to use?

[Adam Hardy]

Adam Hardy on 05/11/10 01:03, wrote:
> I have a small LAN at home connected up to the net via a DSL modem on a gateway 
> machine running lenny and iptables, and newly smokeping to monitor it all.
> 
> I have a problem with some software running on a windows machine on the LAN 
> where the company who wrote the sofware say that there must be something wrong 
> with my internet connection, but can't help any further.
> 
> I'd like to know what tests I can run to verify that my config and my firewall 
> are OK, especially regarding PMTU.

Thanks for the replies.

Essentially the problem is that my client software for a financial brokerage 
application logs repeated disconnects from the data servers and this causes the 
display of charts to fail because it can't fetch the required past data.


The disconnects occur mainly in the late afternoon and evening London time 
(GMT+0) and it makes the application virtually unusable. Magically at 23:00 
every night the disconnects stop and I am able to work normally.

This is why I would like to monitor it in smokeping, but I haven't found a test 
that will fail yet which correlates with the application's disconnect phases.


First of all though I would like to set up a smokeping monitor that shows that 
pathway MTU discovery is working. I thought I could use ping for this, with 
parameters to prevent fragmentation. However since the path is liable to change, 
this is probably tricky.

I am familiar with MTU but not an expert. That wikipedia page was very good and 
had a few points that were totally new to me - for instance that network 
switches have built-in capability to detect when a device is jabbering and block 
it until it resumes proper operation. I wonder whether my connection is 
perceived by a switch out there as 'jabber'. That would be what I hope to monitor.

When I said I dropped my firewall rules, I meant I left the NAT'ing rules in 
there and just had an open gateway. This is what I mean:


Chain PREROUTING (policy ACCEPT 0 packets, 0 bytes)
  pkts bytes target     prot opt in     out     source               destination 


Chain POSTROUTING (policy ACCEPT 1 packets, 60 bytes)
  pkts bytes target     prot opt in     out     source               destination 

     0     0 MASQUERADE  all  --  *      eth1    192.168.0.0/24       0.0.0.0/0 


Chain OUTPUT (policy ACCEPT 1 packets, 60 bytes)
  pkts bytes target     prot opt in     out     source               destination 


Unfortunately the DSL modem I use also has an iptables firewall, undocumented by 
the manufacturer, which I can telnet to drop the rules. On the modem I assume 
that I don't need any NAT'ing at all or anything else. No rules at all, right?

I expect that I will install wireshark to monitor my iptables in action, but I 
think at this point I would rather set up tests to monitor what is going on with 
the internet and the broker client software, around the clock just in case my 
ISP is doing something strange with the broadband connection.

I have thought about getting another ISP and trying out the connectivity there, 
but that could be very dissatisfactory is another type of broadband error crops 
up, and anyway all non-British-Telecom ISPs except cable and satellite use the 
British Telecom equipment closely.

I hope that puts my issue into a more comprehensible light.

Thanks
Adam