perfSONAR User Q&A and Other Discussion

[Daniel Doyle <>]

Hi Mike,

(As an aside it's a lot more useful if you reply-all so that the list can see the discussion and maybe others can chime in with answers, or at least see answers to their own similar issues. Also, then you're not blocked entirely on me necessarily. :) )

You said the MaDDash entries were all done by hostname - are the entries in regular testing also by hostname? Assuming you're on 3.5, you can see this in /opt/perfsonar_ps/regular_testing/etc/regular_testing.conf.

-Dan

On Feb 22, 2016, at 6:01 PM, Michael Reece <> wrote:

I ran the command from a terminal that you suggested above, here is the result:
<image.png>

Does that mean that the test was successful? Both of these machines were recently given the private IP addresses.

On Mon, Feb 22, 2016 at 5:46 PM, Michael Reece <> wrote:

When try to configure a test in the toolkit I see this screen:
<image.png>

On Mon, Feb 22, 2016 at 9:19 AM, Michael Reece <> wrote:

I am not the one who did it, but I believe that they have been
readdressed into RFC1918 addresses. At least they all have 10.* IP
addresses now. I'm not on campus at moment so I cannot run any tests,
but will be later today. I just found it odd that the nodes couldn't
run tests with each other, yet I could SSH into them just fine.

On Mon, Feb 22, 2016 at 8:50 AM, Daniel Doyle <> wrote:
> Hi Mike,
>
> Sorry if this is a dumb question - when you say "they are now all private", does this mean you re-addressed them into RFC1918 addresses? Seems like doing that could have caused some different ACLs to come into play, or otherwise just made them unroutable from one to another if they are in separate private networks. perfSONAR today doesn't really work in a NAT'd environment very well.
>
> If you log into one are you able to initiate tests manually? Something like "bwctl -c $other_private_address" from one of the newly addressed hosts.
>
> -Dan
>
>> On Feb 8, 2016, at 4:36 PM, Michael Reece <> wrote:
>>
>> We had a working perfSONAR mesh setup and the MaDDash and everything
>> was working fine. Some of the nodes were on a public network and my
>> boss wanted them to be private. They are now all private, but all of
>> the nodes changed seem to be unable to run tests with one another. All
>> of the affected nodes have the "Unable to retrieve data" orange color
>> on the MaDDash grid. I've looked into the YAML configuration file for
>> the MaDDash, but whenever it references a node, it does so by hostname
>> not by IP address. I don't understand why switching from a public to
>> private network would affect the nodes. I am able to SSH into all the
>> nodes fine, and also able to access each node's toolkit. Any help
>> would be appreciated.
>>
>> --Mike Reece
>
> Dan Doyle
> GlobalNOC Software Developer
> 1-812-856-3892
>
>
>
>

Dan Doyle

GlobalNOC Software Developer

1-812-856-3892

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail