[lustre-discuss] IO500 SC20 Call for Submission

committee at io500.org committee at io500.org
Thu Oct 1 09:40:00 PDT 2020


CALL FOR IO500 SUBMISSION

Deadline: 30 October 2020 AoE 

Stabilization period: 1st October -- 9th October 2020 AoE 

The IO500 [1] is now accepting and encouraging submissions for the
upcoming 7th IO500 list, to be revealed at the IO500 Virtual BOF during
SC20. Once again, we are also accepting submissions to the 10 Node I/O
Challenge to encourage submission of small scale results. The new ranked
lists will be announced at our Virtual SC20 BoF. We hope to see you, and
your results, there.  

A new change for the upcoming submission procedure is the introduction
of a stabilization period that aims to harden the benchmark. The final
benchmark is released at the end of this period. During the
stabilization we encourage the community to test the proper execution of
the benchmark and provide us with feedback. We will apply bug fixes to
the code base and expect that results obtained will be valid as full
submission. We also continue with another list for the Student Cluster
Competition, since IO500 is used during this competition. 

Also new this year is that we have partnered with Anthony Kougkas' team
at Illinois Institute of Technology to evaluate the submission metadata
describing the storage system on which the test was run to improve the
quality and usefulness of the data IO500 collects. You may be contacted
by one of his students to clarify one or more of the metadata items from
your submission(s). We would appreciate, but do not require, your
cooperation to help improve the submission metadata quality. Results
from their work will be fed back to improve our submission process for
future lists. 

The IO500 benchmark suite is designed to be easy to run, and the
community has multiple active support channels to help with any
questions. Please submit results from your system, and we look forward
to seeing many of you at SC20! Please note that submissions of all sizes
are welcome, including multiple submissions from different storage
systems/tiers at a single site.  The website has customizable sorting so
it is possible to submit on a small system and still get a very good
per-client score, for example. Additionally, the list is about much more
than just the raw rank; all submissions help the community by collecting
and publishing a wider corpus of data. More details below. 

Following the success of the Top500 in collecting and analyzing
historical trends in supercomputer technology and evolution, the IO500
[1] was created in 2017, published its first list at SC17, and has grown
continuously since then. The need for such an initiative has long been
known within High-Performance Computing; however, defining appropriate
benchmarks had long been challenging. Despite this challenge, the
community, after long and spirited discussion, finally reached consensus
on a suite of benchmarks and a metric for resolving the scores into a
single ranking. 

The multi-fold goals of the benchmark suite are as follows: 

 	* Maximizing simplicity in running the benchmark suite
 	* Encouraging complexity in tuning for performance
 	* Allowing submitters to highlight their "hero run" performance
numbers
 	* Forcing submitters to simultaneously report performance for
challenging IO patterns.

Specifically, the benchmark suite includes a hero-run of both IOR and
mdtest configured however possible to maximize performance and establish
an upper-bound for performance. It also includes an IOR and mdtest run
with highly prescribed parameters in an attempt to determine a
lower-bound on the performance. Finally, it includes a namespace search,
as this has been determined to be a highly sought-after feature in HPC
storage systems that have historically not been well-measured.
Submitters are encouraged to share their tuning insights for
publication. 

The goals of the community are also multi-fold: 

 	* Gather historical data for the sake of analysis and to aid
predictions of storage futures
 	* Collect tuning information to share valuable performance
optimizations across the community
 	* Encourage vendors and designers to optimize for workloads beyond
"hero runs"
 	* Establish bounded expectations for users, procurers, and
administrators

10 NODE I/O CHALLENGE

The 10 Node Challenge is conducted using the regular IO500 benchmark,
however, with the rule that exactly 10 client nodes must be used to run
the benchmark. You may use any shared storage with, e.g., any number of
servers. When submitting for the IO500 list, you can opt-in for
"Participate in the 10 compute node challenge only", then we will not
include the results into the ranked list. Other 10-node node submissions
will be included in the full list and in the ranked list. We will
announce the result in a separate derived list and in the full list but
not on the ranked IO500 list at https://io500.org/ [2]  

BIRDS-OF-A-FEATHER

Once again, we encourage you to submit [1], to join our community, and
to attend our virtual BoF "The IO500 and the Virtual Institute of I/O"
at SC20, where we will announce the new IO500 list, the 10 node
challenge list, and the Student Cluster Competition list. We look
forward to answering any questions or concerns you might have. 

 	* [1] http://www.vi4io.org/io500/submission [3] 

Thanks, 

The IO500 Committee <committee at io500.org> 

 

Links:
------
[1] http://io500.org/
[2] https://io500.org/
[3] http://www.vi4io.org/io500/submission
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20201001/7611b021/attachment.html>


More information about the lustre-discuss mailing list