perfSONAR User Q&A and Other Discussion

[Andrew Gallo <>]

Yes, this is coming from esmond.  The URL I'm using looks something like
> http://127.0.0.1/esmond/perfsonar/archive/<id>/histogram-owdelay/base/0?time-start=1712577600&time-end=1713182400&limit=10000


This is helpful.  Thanks for the response.




On 4/19/2024 2:54 PM, Andrew Lake wrote:
> Hi Andrew,
>
> I assume this is coming out of the old esmond interface? You are correct
> that the key is the bucket and the value is the count of results belonging
> to that. IIRC The buckets are always normalized to milliseconds so in your
> example 198 samples measured at -.01ms, 95 at -.02ms and so forth. The
> bucket resolution is dependent on what’s configured for the test using the
> “bucket-width” parameter. I believe 1/100th of a millisecond is the default
> if not specified. If it is set to something else you’ll see it in the
> “bucket-width” parameter of the metadata for the test. You are also correct
> that buckets with no observed values are suppressed.
>
> Thanks,
> Andy
>
> On April 18, 2024 at 11:55:37 AM, Andrew Gallo ()
> wrote:
>
> I have a couple of questions about the format of the results returned
> from the API for owamp tests.
>
>
> Here is a sample I'm looking at
>
> > {
> > "ts": 1710482312,
> > "val": {
> > "-0.01": 198,
> > "-0.02": 95,
> > "-0.03": 28,
> > "-0.04": 5,
> > "0.0": 156,
> > "0.01": 82,
> > "0.02": 28,
> > "0.04": 7,
> > "0.07": 1
> > }
> > },
>
> The structure looks like key:value pairs, with the keys being the
> latency measurements and the values being the number of results in that
> 'bucket.'
>
> Is this correct?
>
> Can I assume...
> 1. The resolution (or ‘bucket size’) is always 0.01 msec?
> 2. Buckets with no results are suppressed?
>
>
>
> Thank you

--
________________________________
Andrew Gallo
The George Washington University

Attachment: OpenPGP_signature.asc
Description: OpenPGP digital signature