perfSONAR User Q&A and Other Discussion

[Andrew Lake <>]

On Thu, Oct 1, 2015 at 9:48 AM, Brian Candler <> wrote:

One of my perfsonar nodes has upgraded itself to 3.5. The graphical interface is pretty but I'd like to report the following things which no longer work.

(I can send the actual server links off-list to developers; these are non-public servers)

Many of the breakages relate to IPv6.

The "tests result" section still shows two rows, one for v4 and one for v6:

cov-client-1.snipped    cov-perf-1.snipped
X.X.X.X                 Y.Y.Y.Y
Details | Traceroute

cov-client-1.snipped    cov-perf-1.snipped
XXXX:XXXX:XXXX:XXXX     YYYY:YYYY:YYYY:YYYY
Details

Problem 1: clicking on either of the two "Details" links shows the same graph. The graph headers say

         Source                       Destination
cov-client-1.snipped              cov-perf-1.snipped
X.X.X.X, XXXX:XXXX:XXXX:XXXX Y.Y.Y.Y,YYYY:YYYY:YYYY:YYYY

Previously, I could get different graphs for the v4 and v6. Now I can only get one, and it's not even clear which it's showing.

Clicking "link to this chart" gives a URL which contains
graphWidget.cgi?source=cov-client-1.snipped&dest=cov-perf-1.snipped

whereas before it would have given either the IPv4 or IPv6 addresses.

As you more or less diagnosed, this sounds like a bug where the hostname is getting passed instead of the IP. I haven't looked close enough to confirm locally, but if that's the case it should hopefully be a quick fix and we'll push out a new RPM for it when it's ready. 

(2) From the time of the upgrade, I get red dots showing:

"04:57:41 10/01/2015
Error from bwctl/traceroute:
Problem parsing traceroute output: Can't find hop number in output: traceroute6 to XXXX:XXXX:0:2::254 (XXXX:XXXX:0:2::254) from XXXX:XXXX:1:64::252, 64 hops max, 12 byte packets at /usr/share/perl5/vendor_perl/Net/Traceroute.pm line 624.
04:57:41 10/01/2015"

This might also explain why there is no "traceroute" link in the "test results" section for the IPv6 addresses.

Hmm that's strange, if you run the traceroute6 command in question by hand, anything of note about the output? The traceroute parsing hasn't changed in 3.5 as far as I can tell but it's a little tricky because we rely on the Net::Traceroute module to do the parsing. 

 

(3) From the time of the upgrade, ping tests have stopped working (the dotted yellow line for "reverse ping" has just stopped)

But Configuration > Tests still shows the section (which I called "Local ping test"), and I can see the test is still there.

Do you have bwctl running on the hosts running the ping tests? Someone else reported a similar issue and they somehow had been getting results in 3.4 without bwctl enabled (which should not have been possible since we use bwping for all ping tests starting in 3.4). 

 

(4) Probably unrelated, but click on services directory and it takes you to
http://stats.es.net/ServicesDirectory/
there's almost nothing (just a few red dots in Asia)

Yeah, that's unrelated to the upgrade. It's been showing ~1500 hosts for me all morning. The page has been a little flaky though, the LS infrastructure seems to be fine, so it's just GUI issues. 

 

Regards,

Brian.