perfSONAR User Q&A and Other Discussion

[Jason Zurawski <>]

Understanding the differences in
configuration/performance/behavior/design of “MEB-edge —> MEB-Core”
and “CSRB-edge —> CSRB-Core” would be where I would focus.
Specifically:

- Are the results consistently repeatable (e.g. systematic low
performance) or do they vary (congestion/traffic patterns)?

- Are there any red flags being thrown by devices in the path (error
counters, etc.)

- Can you replicate the same behavior to other externally located
devices besides the XSEDE node

- How does latency relationship to TCP tests play into this (e.g. to
your point, testing < 5ms between the nodes isn’t useful, test to
things further out)

Thanks;

-jason

On January 28, 2019 at 11:19:31 AM, Qayoom, Mohamad A.
() wrote:
> Hi Jason,
>
> I won’t be able to draw a diagram for a couple of hours, but here are the 
> network paths:
>
>
>
> LSUHSC-pS-MEB —> MEB-edge —> MEB-Core —> RCB-Core —> firewall —> RCB-Border 
> —> LONI
> —>>>
>
> LSUHSC-pS-CSRB —> CSRB-edge —> CSRB-Core —> RCB-Core —> firewall —> 
> RCB-Border —> LONI
> —>>>
>
> Both nodes are ACLes off.
>
>
> Thank you,
> Mohamad Qayoom
>
> > On Jan 28, 2019, at 12:48 PM, Jason Zurawski wrote:
> >
> > *EXTERNAL EMAIL: EVALUATE*
> >
> > “What’s in the path” - more accurately, what is different about the
> > path. Would be really helpful to see network diagrams of how the
> > paths between your internal nodes and the external node differ?
> >
> > I can’t seem to get a test to either node to work - are these ACLed off?
> >
> > Thanks;
> >
> > -jason
> >
> > On January 28, 2019 at 10:30:43 AM, Qayoom, Mohamad A.
> > () wrote:
> >> Hello,
> >>
> >> I am comparing two internal perfSONAR nodes and I am seeing lower 
> >> throughput on one
> of
> >> the two nodes. The internal nodes are lsuhsc-ps-meb and lsuhsc-ps-csrb.
> >>
> >>
> >> Lower Throughput
> >>
> >> [file:///Users/zurawski/Library/Group 
> >> Containers/2E337YPCZY.airmail/Library/Application
> ]
> >>
> >>
> >> Expected Throughput
> >>
> >> [file:///Users/zurawski/Library/Group 
> >> Containers/2E337YPCZY.airmail/Library/Application
> ]
> >>
> >>
> >> The throughput between the two internal nodes is fine. I have run a few 
> >> pScheduler tests,
> >> but I couldn't isolate where the issues is. Can someone please suggest 
> >> any further
> tests
> >> to isolate the problem?
> >>
> >>
> >> Thank you,
> >>
> >> Mohamad Qayoom
> >> Department of Information Technology
> >> LSU Health New Orleans
> >> Office: (504) 568-5232
> >> Blackberry: (504) 234-4853
> >> 
> >>
> >>
> >>
> >> --
> >> To unsubscribe from this list: 
> >> https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.internet2.edu%2Fsympa%2Fsignoff%2Fperfsonar-user&data=02%7C01%7Cmqayoo%40lsuhsc.edu%7Ccf3943d6efa749cd2e4808d6855128db%7C3406368982d44e89a3281ab79cc58d9d%7C0%7C0%7C636842980973696135&sdata=lACVHjNG3UFLwmqg4ewE9V9PIdSeNbljpbpJozcd1xs%3D&reserved=0
> >>
> --
> To unsubscribe from this list: 
> https://lists.internet2.edu/sympa/signoff/perfsonar-user
>