ROMIO

We wrapped up another SC week in Denver .

• Saw a good bit of interest in ROMIO features, performance, and supported drivers at our I/O tutorial Parallel I/O in Practice
• I’m excited about some of the new features discussed at the HDF5 BoF, including an “everyone reads the same data” optimization that was a tremendous pain to implement in ROMIO, but (with a bit of additional communication) easy to implement at the HDF5 layer

presentations

If you would like to learn more about I/O in High Performance Computing, come check out our SC 2017 tutorial Parallel I/O in Practice. We will cover the hardware and software that makes up the software stack on large parallel computers. MPI-IO takes up a big chunk of time, as do the I/O libraries which typically sit on top of MPI-IO.
If you are interested in ROMIO, you hopefully are familiar with Darshan for collecting statistics and otherwise characterizing you your I/O patterns. These other Darshan-related events will likely be of interest to you:
Thursday’s “Analyzing Parallel I/O” BoF by Phil Carns and Julian Kunkel.  Several guests will share their experiences in analyzing I/O performance and participate in a panel discussion:
http://sc17.supercomputing.org/presentation/?id=bof136&sess=sess364
https://wr.informatik.uni-hamburg.de/events/2017/bof-analyzing
Glenn Lockwood will be presenting a paper entitled “UMAMI: A Recipe for Generating Meaningful Metrics through Holistic I/O Performance Analysis” at the PDSW-DISCS workshop on Monday as well!  It highlights how to integrate I/O monitoring data from multiple sources into a coherent view.
http://www.pdsw.org/index.shtml

presentations

Rajeev Thakur, original author of ROMIO, will be speaking at the “Celebrating 25 years of MPI” symposium on 25 September.   It’s one of the workshops held in conjunciton with EuroMPI 2017.  If you are traveling to Argonne for the conference, stop by and let him know what you think about ROMIO.
For more information, visit http://www.mcs.anl.gov/mpi-symposium/
 

presentations

If you would like to learn more about I/O in High Performance Computing, come check out our SC 2016 tutorial Parallel I/O in Practice. We will cover the hardware and software that makes up the software stack on large parallel computers. MPI-IO takes up a big chunk of time, as do the I/O libraries which typically sit on top of MPI-IO.

presentations

SC 2015 is just a few days away. ROMIO will be part of our Sunday I/O tutorial. “Parallel I/O In Practice”: our 11th full-day tutorial on Parallel I/O hardware, software, tools and strategies. ROMIO, Parallel-NetCDF, HDF5, and Darshan will be the focus of the afternoon half of the day. File systems and storage hardware will be the focus of the morning.
You will also be able to find Rob Latham wandering around Austin: he would crawl over a mountain of vendor trinkets for a conversation about MPI-IO. Don’t be shy!

presentations

Brock Palen and Jeff Squyres chatted with Rajeev Thakur and Rob Latham about ROMIO for the RCE Podcast. That discussion was episode 66.
We talked about ROMIO’s history, how ROMIO serves as the MPI-IO implementation for several MPI implementations, file system support, and what’s coming down the road.

presentations