Ssio-2018
Threads by month
- ----- 2026 -----
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2025 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2024 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2023 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2022 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2021 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2020 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2019 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
- ----- 2018 -----
- December
- November
- October
- September
- August
- July
- June
- May
- April
- March
- February
- January
April 2018
- 5 participants
- 8 discussions
Has there been any movement on the virtual summit tentatively scheduled for
next week?
Glenn
On Mon, Apr 9, 2018 at 11:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
> Hi all,
>
> Thanks for taking the time to fill out the polls. While we don't have the
> go-ahead for the virtual summer or workshop yet, let's hold the following
> dates:
> - May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans,
> contact info TBD)
> - August 1-2 for a tentative workshop (probably in DC/MD)
>
> Thanks!
>
> Rob
>
>
> On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
>
> Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for
> getting one out of two?). Please fill out the Doodle polls when you can. --
> Rob
>
> On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
>
> Hi all,
>
> Scott asked a good question that I will summarize as "what is going on?".
> Here's what I know:
>
> *Virtual Summit. *Lucy has asked for permission to have our virtual
> summit, back on the 21st. I will double-check with her on status of that
> request. Meanwhile we should all gather our thoughts for the summit:
> - Think about what you think the big issues are in SSIO research for
> enabling DOE science, talk with your team and colleagues.
> - Fill out this poll; we're optimistic that we will have permission to do
> this soon:
> https://doodle.com/poll/mvfxb8fcqucimtim
>
> *Scope.* We've been asked to think about the data and learning
> applications in addition to traditional simulation. I don't think this is a
> surprise, but it does mean that we need to think about those communities as
> we think about who to include in the discussion/workshop and what to have
> in our initial report.
>
> *Organizers. *Gary has asked to take a step back but is happy to chime in
> as needed. Here's our revised team:
> - Brad Settlemyer (LANL)
> - Scott Klasky (ORNL)
> - Glenn Lockwood (LBNL)
> - Kathryn Mohror (LLNL)
> - Lee Ward (SNL)
>
> I think everyone probably knows Brad. I'm sure he'll do a great job. Gary
> will remain on the mailing list, too.
>
> *Meeting Dates.* Lucy and I have chatted on possible dates, and we want
> to get some dates on the calendar. We're looking for two days in
> July/August. Here's a short set of dates that work for me and Lucy; we're
> hoping to get lucky and get a hit. Note that we didn't include Mon. or Fri.
> in these so people can travel during the week:
> - https://doodle.com/poll/nw4xsqpx2agd3dhh
>
> *Attendees.* We should start thinking about attendees. I've attached the
> list of people from the SSIO workshop back in late 2014 as a starting
> point. Can someone volunteer to create a Google spreadsheet that we could
> use for organizing possible participants. Note that we don't have
> permission to have the meeting at this point, so we're really just
> discussing who we think key thinkers are in the area.
>
> I'm also attaching the EH and ML pre-workshop documents as examples of the
> kind of report we need to prepare.
>
> Thanks,
>
> Rob
> ---
>
> On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV> wrote:
>
> Rob tasked me with trying to kick this off. I had something different in
> mind but I suspect a call is a better idea. Thanks!
>
> Sent from my iPhone
>
> On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov> wrote:
>
> Hi Rob,
> Has there been a doodle poll for this?
>
> Scott
>
> *From:* ssio-2018 [mailto:[email protected]
> <ssio-2018-bounces(a)lists.mcs.anl.gov>] *On Behalf Of *Ross, Robert B.
> *Sent:* Friday, January 12, 2018 1:18 PM
> *To:* ssio-2018(a)lists.mcs.anl.gov
> *Subject:* Re: [Ssio-2018] SSIO 2018 Workshop Planning
>
> Forgot the attachment, of course. -- Rob
>
>
>
> On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
>
> Hi all,
>
> I've chatted with all of you previously about this, but I've been lax in
> getting moving. Lucy has asked me to organize a workshop to help define
> priority research directions for scalable storage and I/O (SSIO), ideally
> leading to a call for proposals some time in 2018. Additionally, there is
> some interest from Almadena on the NSF side to restart an inter-agency
> activity in this space. We will be the organizers that drive the ASCR
> process forward. We'll see also how things develop on the inter-agency
> front and what role we might play in that process.
>
> *Our goal is to organize our thoughts about future research directions,
> produce an initial document capturing that thinking and describing state of
> the art, plan and hold a workshop to identify a set of priority research
> directions, and produce a final workshop report capturing this information.*
>
> My initial plan is based off the experiences from the old HEC-FSIO
> process, the more recent 2014 ASCR SSIO workshop (for which there is a 2015
> report), and the most recent Extreme Heterogeneity workshop planning (for
> which we will have the workshop soon). All of this is up for
> discussion/negotiation.
>
> *Summary and action items.*
> - Think about what you think the big issues are in SSIO research for
> enabling DOE science, talk with your team and colleagues.
> - Start putting together the "virtual" summit materials.
> - Look for the Doodle poll for a summit date.
> - Look for a Doodle poll for a time we could regularly meet to discuss
> progress.
>
> Details follow. Thanks in advance for your help in making this a success!
>
> Rob
> ---
>
> *Organizers. *Here's our team:
> - Gary Grider (LANL)
> - Scott Klasky (ORNL)
> - Glenn Lockwood (LBNL)
> - Kathryn Mohror (LLNL)
> - Lee Ward (SNL)
>
> Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be
> included and possibly take over as contact before we are done.
>
> We have a mailing list: ssio-2018(a)lists.mcs.anl.gov for discussion. You
> have been added to the list. Please let me know ASAP if for some reason you
> cannot help.
>
>
> *Topical areas and initial discussion. *Five focus areas were discussed
> for further investigation at the prior workshop.
> - HW/SW architectures
> - Metadata, name spaces, and provenance
> - Supporting science data
> - Integrating with external services
> - Understanding SSIO systems
>
> I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th
> of a day, rather than flying somewhere for a day to meet in person. Each
> organizer would present their view on challenges as relate to mission and
> discuss if/how they think we should adjust the topical areas in response to
> changes since 2015. I think of these as 20 minute presentations that are
> devoid of our own research specifics/propaganda and focus on these big
> picture items. From this we would select a set of topical areas used to
> organize the factual status document (see below).
>
>
> *Factual status document. *We are modeling our process after the BES
> "basic research needs" approach (see attachment), with adaptations to the
> scope of the area and ASCR needs. This process includes the generation of a
> factual status document (FSD) prior to the workshop. There is a separate
> group working down this path for an Extreme Heterogeneity FSD and workshop,
> too. The FSD can be thought of as the background sections of the final
> workshop document, and that's how we will approach it.
>
> With this in mind, we will organize the document around our initial
> guesses at topical areas, understanding that the topical areas may have to
> be adjusted in the final report as a result of discussion. The goal would
> be to complete the following for the FSD:
>
> - Introduction
> - Science Challenges
> - Technology Challenges
> - Knowledge Gaps -- what we need to know that we don't know
> - Research Needs
> - Topic 1
> - Definition/summary of area
> - Current status
> - Topic 2 ...
> - Appendix: References
>
> This is likely to be a 50 page document.
>
> This document would be partitioned into sections to be managed by
> individual organizers. Rather than requesting white papers from potential
> attendees, organizers would invite experts in respective areas to
> participate in the development of requisite sections. It might be easiest
> to develop this document using an online system such as Google Docs,
> understanding that eventually it would be turned into a Word document and
> finally into a PDF for publication as the workshop report.
>
>
> *Position papers.* In the Extreme Heterogeneity planning, we asked for
> position papers. We got over 100 of them. Parsing them has been a real
> challenge for the organizers. We should think about what approach we might
> employ for gathering input from the community prior to the workshop and
> what we would do with that material prior to the workshop vs. incorporating
> into the final report.
>
>
> *Workshop. *We will shoot for around 60 attendees, maybe 30/20/10
> lab/university/industry. This helps keep the justification and planning
> fairly simple. We would plan for a 1.75-day workshop to facilitate covering
> five topical areas as separate, single-track discussions. If for some
> reason we end up needing only 1.5 days for the meeting proper, this would
> leave more time for the organizers to meet and discuss report completion.
>
> We will (mostly) assume that attendees have read the FSD prior to arrival
> at the meeting. Never the less, we will spend 1/2 day bringing the
> attendees up to speed, focusing on the background material from the FSD.
> This is similar to the opening plenary session portion of the BES workshop
> structure.
>
> The remainder of the meeting could follow the model from the 2014 SSIO
> meeting: 90-minute moderated discussion sessions, with notes captured live
> in PPT. With 1.75 days we can accommodate 5 total discussion topics. We
> will place a topic of broad interest last to encourage attendees to
> participate for the full meeting. These sessions will identify current
> status in the area, its CS challenges, its opportunities, and its research
> needs, reflecting the breakout session portion of the BES workshop
> structure. Because the meeting is single track, there is no need for
> outbrief sessions (i.e., the participants will have been present for all
> discussions).
>
>
> *Report. *The final report will be generated by augmenting the FSD with
> material summarizing the outcomes of the workshop. We will target a 75-100
> page document (the 2015 report was approximately 90 pages). The outline for
> the report will look something like this:
>
> - Executive Summary
> - Introduction
> - Science Challenges
> - Technology Challenges
> - Knowledge Gaps (need to understand what this typically includes)
> - Research Needs
> - Topic 1
> - Definition/summary of area
> - Current status
> - CS challenges and opportunities
> - Priority research directions (3-5)
> - Topic 2 ...
> - Priority research directions
>
> _______________________________________________
> ssio-2018 mailing list
> ssio-2018(a)lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
>
>
>
> _______________________________________________
> ssio-2018 mailing list
> ssio-2018(a)lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
>
>
> <2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-
> doc-20180124.pdf>
>
>
>
>
> _______________________________________________
> ssio-2018 mailing list
> ssio-2018(a)lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
>
>
1
0
Can rename the column. -- Rob
On Apr 9, 2018, at 6:32 PM, Mohror, Kathryn <mohror1(a)llnl.gov<mailto:[email protected]>> wrote:
Hi Scott,
I think I/O fits under ‘Storage Systems’.
Kathryn
From: Klasky, Scott A. <klasky(a)ornl.gov<mailto:[email protected]>>
Sent: Monday, April 9, 2018 4:27 PM
To: Mohror, Kathryn <mohror1(a)llnl.gov<mailto:[email protected]>>; Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>>; ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Why don't we have an io column? I vote to have this column
Scott
________________________________
From: Mohror, Kathryn <mohror1(a)llnl.gov<mailto:[email protected]>>
Date: April 9, 2018 at 11:54:47 PM GMT+2
To: Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>>, ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> <ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi everybody,
I checked the spreadsheet of potential attendees and no one has entered any yet!
Rob, are we including only DOE folks? Or are university and industry people included as well?
Please take a bit of time to start to jot down some names.
The spreadsheet is here: https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxzCg…
Kathryn
From: ssio-2018 <ssio-2018-bounces(a)lists.mcs.anl.gov<mailto:[email protected]>> On Behalf Of Ross, Robert B.
Sent: Monday, April 9, 2018 11:33 AM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi all,
Thanks for taking the time to fill out the polls. While we don't have the go-ahead for the virtual summer or workshop yet, let's hold the following dates:
- May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans, contact info TBD)
- August 1-2 for a tentative workshop (probably in DC/MD)
Thanks!
Rob
On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for getting one out of two?). Please fill out the Doodle polls when you can. -- Rob
On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
Scott asked a good question that I will summarize as "what is going on?". Here's what I know:
Virtual Summit. Lucy has asked for permission to have our virtual summit, back on the 21st. I will double-check with her on status of that request. Meanwhile we should all gather our thoughts for the summit:
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Fill out this poll; we're optimistic that we will have permission to do this soon:
https://doodle.com/poll/mvfxb8fcqucimtim
Scope. We've been asked to think about the data and learning applications in addition to traditional simulation. I don't think this is a surprise, but it does mean that we need to think about those communities as we think about who to include in the discussion/workshop and what to have in our initial report.
Organizers. Gary has asked to take a step back but is happy to chime in as needed. Here's our revised team:
- Brad Settlemyer (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
I think everyone probably knows Brad. I'm sure he'll do a great job. Gary will remain on the mailing list, too.
Meeting Dates. Lucy and I have chatted on possible dates, and we want to get some dates on the calendar. We're looking for two days in July/August. Here's a short set of dates that work for me and Lucy; we're hoping to get lucky and get a hit. Note that we didn't include Mon. or Fri. in these so people can travel during the week:
- https://doodle.com/poll/nw4xsqpx2agd3dhh
Attendees. We should start thinking about attendees. I've attached the list of people from the SSIO workshop back in late 2014 as a starting point. Can someone volunteer to create a Google spreadsheet that we could use for organizing possible participants. Note that we don't have permission to have the meeting at this point, so we're really just discussing who we think key thinkers are in the area.
I'm also attaching the EH and ML pre-workshop documents as examples of the kind of report we need to prepare.
Thanks,
Rob
---
On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV<mailto:[email protected]>> wrote:
Rob tasked me with trying to kick this off. I had something different in mind but I suspect a call is a better idea. Thanks!
Sent from my iPhone
On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov<mailto:[email protected]>> wrote:
Hi Rob,
Has there been a doodle poll for this?
Scott
From: ssio-2018 [mailto:[email protected]] On Behalf Of Ross, Robert B.
Sent: Friday, January 12, 2018 1:18 PM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Forgot the attachment, of course. -- Rob
On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
I've chatted with all of you previously about this, but I've been lax in getting moving. Lucy has asked me to organize a workshop to help define priority research directions for scalable storage and I/O (SSIO), ideally leading to a call for proposals some time in 2018. Additionally, there is some interest from Almadena on the NSF side to restart an inter-agency activity in this space. We will be the organizers that drive the ASCR process forward. We'll see also how things develop on the inter-agency front and what role we might play in that process.
Our goal is to organize our thoughts about future research directions, produce an initial document capturing that thinking and describing state of the art, plan and hold a workshop to identify a set of priority research directions, and produce a final workshop report capturing this information.
My initial plan is based off the experiences from the old HEC-FSIO process, the more recent 2014 ASCR SSIO workshop (for which there is a 2015 report), and the most recent Extreme Heterogeneity workshop planning (for which we will have the workshop soon). All of this is up for discussion/negotiation.
Summary and action items.
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Start putting together the "virtual" summit materials.
- Look for the Doodle poll for a summit date.
- Look for a Doodle poll for a time we could regularly meet to discuss progress.
Details follow. Thanks in advance for your help in making this a success!
Rob
---
Organizers. Here's our team:
- Gary Grider (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be included and possibly take over as contact before we are done.
We have a mailing list: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> for discussion. You have been added to the list. Please let me know ASAP if for some reason you cannot help.
Topical areas and initial discussion. Five focus areas were discussed for further investigation at the prior workshop.
- HW/SW architectures
- Metadata, name spaces, and provenance
- Supporting science data
- Integrating with external services
- Understanding SSIO systems
I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th of a day, rather than flying somewhere for a day to meet in person. Each organizer would present their view on challenges as relate to mission and discuss if/how they think we should adjust the topical areas in response to changes since 2015. I think of these as 20 minute presentations that are devoid of our own research specifics/propaganda and focus on these big picture items. From this we would select a set of topical areas used to organize the factual status document (see below).
Factual status document. We are modeling our process after the BES "basic research needs" approach (see attachment), with adaptations to the scope of the area and ASCR needs. This process includes the generation of a factual status document (FSD) prior to the workshop. There is a separate group working down this path for an Extreme Heterogeneity FSD and workshop, too. The FSD can be thought of as the background sections of the final workshop document, and that's how we will approach it.
With this in mind, we will organize the document around our initial guesses at topical areas, understanding that the topical areas may have to be adjusted in the final report as a result of discussion. The goal would be to complete the following for the FSD:
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps -- what we need to know that we don't know
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- Topic 2 ...
- Appendix: References
This is likely to be a 50 page document.
This document would be partitioned into sections to be managed by individual organizers. Rather than requesting white papers from potential attendees, organizers would invite experts in respective areas to participate in the development of requisite sections. It might be easiest to develop this document using an online system such as Google Docs, understanding that eventually it would be turned into a Word document and finally into a PDF for publication as the workshop report.
Position papers. In the Extreme Heterogeneity planning, we asked for position papers. We got over 100 of them. Parsing them has been a real challenge for the organizers. We should think about what approach we might employ for gathering input from the community prior to the workshop and what we would do with that material prior to the workshop vs. incorporating into the final report.
Workshop. We will shoot for around 60 attendees, maybe 30/20/10 lab/university/industry. This helps keep the justification and planning fairly simple. We would plan for a 1.75-day workshop to facilitate covering five topical areas as separate, single-track discussions. If for some reason we end up needing only 1.5 days for the meeting proper, this would leave more time for the organizers to meet and discuss report completion.
We will (mostly) assume that attendees have read the FSD prior to arrival at the meeting. Never the less, we will spend 1/2 day bringing the attendees up to speed, focusing on the background material from the FSD. This is similar to the opening plenary session portion of the BES workshop structure.
The remainder of the meeting could follow the model from the 2014 SSIO meeting: 90-minute moderated discussion sessions, with notes captured live in PPT. With 1.75 days we can accommodate 5 total discussion topics. We will place a topic of broad interest last to encourage attendees to participate for the full meeting. These sessions will identify current status in the area, its CS challenges, its opportunities, and its research needs, reflecting the breakout session portion of the BES workshop structure. Because the meeting is single track, there is no need for outbrief sessions (i.e., the participants will have been present for all discussions).
Report. The final report will be generated by augmenting the FSD with material summarizing the outcomes of the workshop. We will target a 75-100 page document (the 2015 report was approximately 90 pages). The outline for the report will look something like this:
- Executive Summary
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps (need to understand what this typically includes)
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- CS challenges and opportunities
- Priority research directions (3-5)
- Topic 2 ...
- Priority research directions
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
<2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-doc-20180124.pdf>
1
0
Yes, if we want to get traction within ASCR we need to make sure that we have incorporated facility concerns. Inviting some representatives to the meeting is a good step in that direction. Dave might be a good industry representative, too. -- Rob
On Apr 10, 2018, at 10:15 AM, Settlemyer, Bradley Wade <bws(a)lanl.gov<mailto:[email protected]>> wrote:
It might be nice to make sure we cover/count the storage facilities. I
actually think it will be covered without much challenge, but someone like
Dave Dillow might be a nice add in that he used to work for a DOE
facility, and now works for a large cloud facility.
Just a thought.
Cheers,
Brad
On 4/9/18, 5:32 PM, "ssio-2018 on behalf of Mohror, Kathryn"
<ssio-2018-bounces(a)lists.mcs.anl.gov<mailto:[email protected]> on behalf of mohror1(a)llnl.gov<mailto:[email protected]>> wrote:
Hi Scott,
I think I/O fits under ‘Storage Systems’.
Kathryn
From: Klasky, Scott A. <klasky(a)ornl.gov<mailto:[email protected]>>
Sent: Monday, April 9, 2018 4:27 PM
To: Mohror, Kathryn <mohror1(a)llnl.gov<mailto:[email protected]>>; Ross, Robert B.
<rross(a)mcs.anl.gov<mailto:[email protected]>>; ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Why don't we have an io column? I vote to have this column
Scott
________________________________________
From: Mohror, Kathryn <mohror1(a)llnl.gov<mailto:[email protected]>>
Date: April 9, 2018 at 11:54:47 PM GMT+2
To: Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>>,
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> <ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi everybody,
I checked the spreadsheet of potential attendees and no one has entered
any yet!
Rob, are we including only DOE folks? Or are university and industry
people included as well?
Please take a bit of time to start to jot down some names.
The spreadsheet is here:
https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxzC
gRjPj9YEY/edit?usp=sharing
<https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxz
CgRjPj9YEY/edit?usp=sharing>
Kathryn
From: ssio-2018 <ssio-2018-bounces(a)lists.mcs.anl.gov<mailto:[email protected]>>
On Behalf Of Ross, Robert B.
Sent: Monday, April 9, 2018 11:33 AM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi all,
Thanks for taking the time to fill out the polls. While we don't have the
go-ahead for the virtual summer or workshop yet, let's hold the following
dates:
- May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans,
contact info TBD)
- August 1-2 for a tentative workshop (probably in DC/MD)
Thanks!
Rob
On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for
getting one out of two?). Please fill out the Doodle polls when you can.
-- Rob
On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
Scott asked a good question that I will summarize as "what is going on?".
Here's what I know:
Virtual Summit. Lucy has asked for permission to have our virtual summit,
back on the 21st. I will double-check with her on status of that request.
Meanwhile we should
all gather our thoughts for the summit:
- Think about what you think the big issues are in SSIO research for
enabling DOE science, talk with your team and colleagues.
- Fill out this poll; we're optimistic that we will have permission to do
this soon:
https://doodle.com/poll/mvfxb8fcqucimtim
<https://doodle.com/poll/mvfxb8fcqucimtim>
Scope. We've been asked to think about the data and learning applications
in addition to traditional simulation. I don't think this is a surprise,
but it does mean
that we need to think about those communities as we think about who to
include in the discussion/workshop and what to have in our initial report.
Organizers. Gary has asked to take a step back but is happy to chime in
as needed. Here's our revised team:
- Brad Settlemyer (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
I think everyone probably knows Brad. I'm sure he'll do a great job. Gary
will remain on the mailing list, too.
Meeting Dates. Lucy and I have chatted on possible dates, and we want to
get some dates on the calendar. We're looking for two days in
July/August. Here's a short
set of dates that work for me and Lucy; we're hoping to get lucky and
get a hit. Note that we didn't include Mon. or Fri. in these so people
can travel during the week:
-
https://doodle.com/poll/nw4xsqpx2agd3dhh
<https://doodle.com/poll/nw4xsqpx2agd3dhh>
Attendees. We should start thinking about attendees. I've attached the
list of people from the SSIO workshop back in late 2014 as a starting
point. Can someone volunteer
to create a Google spreadsheet that we could use for organizing possible
participants. Note that we don't have permission to have the meeting at
this point, so we're really just discussing who we think key thinkers are
in the area.
I'm also attaching the EH and ML pre-workshop documents as examples of
the kind of report we need to prepare.
Thanks,
Rob
---
On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV> wrote:
Rob tasked me with trying to kick this off. I had something different in
mind but I suspect a call is a better idea. Thanks!
Sent from my iPhone
On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov> wrote:
Hi Rob,
Has there been a doodle poll for this?
Scott
From: ssio-2018 [mailto:[email protected]] On
Behalf Of Ross, Robert B.
Sent: Friday, January 12, 2018 1:18 PM
To: ssio-2018(a)lists.mcs.anl.gov
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Forgot the attachment, of course. -- Rob
On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
Hi all,
I've chatted with all of you previously about this, but I've been lax in
getting moving. Lucy has asked me to organize a workshop to help define
priority research directions for scalable storage and I/O (SSIO),
ideally leading to a call for proposals some time in 2018. Additionally,
there is some interest from Almadena on the NSF side to restart an
inter-agency activity in this space. We will be the organizers that drive
the ASCR process forward. We'll see also how
things develop on the inter-agency front and what role we might play in
that process.
Our goal is to organize our thoughts about future research directions,
produce an initial document capturing that thinking and describing state
of the art, plan and hold a workshop to identify a set of priority
research directions, and produce a final workshop report capturing this
information.
My initial plan is based off the experiences from the old HEC-FSIO
process, the more recent 2014 ASCR SSIO workshop (for which there is a
2015 report), and the most recent Extreme Heterogeneity workshop planning
(for which we will have the workshop soon). All of this is up for
discussion/negotiation.
Summary and action items.
- Think about what you think the big issues are in SSIO research for
enabling DOE science, talk with your team and colleagues.
- Start putting together the "virtual" summit materials.
- Look for the Doodle poll for a summit date.
- Look for a Doodle poll for a time we could regularly meet to discuss
progress.
Details follow. Thanks in advance for your help in making this a success!
Rob
---
Organizers. Here's our team:
- Gary Grider (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be
included and possibly take over as contact before we are done.
We have a mailing list: ssio-2018(a)lists.mcs.anl.gov for
discussion. You have been added to the list. Please let me know ASAP if
for some reason you cannot help.
Topical areas and initial discussion. Five focus areas were discussed for
further investigation at the prior workshop.
- HW/SW architectures
- Metadata, name spaces, and provenance
- Supporting science data
- Integrating with external services
- Understanding SSIO systems
I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th
of a day, rather than flying somewhere for a day to meet in person. Each
organizer would present their view on challenges as relate
to mission and discuss if/how they think we should adjust the topical
areas in response to changes since 2015. I think of these as 20 minute
presentations that are devoid of our own research specifics/propaganda
and focus on these big picture items. From this
we would select a set of topical areas used to organize the factual
status document (see below).
Factual status document. We are modeling our process after the BES "basic
research needs" approach (see attachment), with
adaptations to the scope of the area and ASCR needs. This process
includes the generation of a factual status document (FSD) prior to the
workshop. There is a separate group working down this path for an Extreme
Heterogeneity FSD and workshop, too. The FSD
can be thought of as the background sections of the final workshop
document, and that's how we will approach it.
With this in mind, we will organize the document around our initial
guesses at topical areas, understanding that the topical areas may have
to be adjusted in the final report as a result of discussion. The
goal would be to complete the following for the FSD:
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps -- what we need to know that we don't know
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- Topic 2 ...
- Appendix: References
This is likely to be a 50 page document.
This document would be partitioned into sections to be managed by
individual organizers. Rather than requesting white papers from potential
attendees, organizers would invite experts in respective areas to
participate in the development of requisite sections. It might be
easiest to develop this document using an online system such as Google
Docs, understanding that eventually it would be turned into a Word
document and finally into a PDF for publication as the
workshop report.
Position papers. In the Extreme Heterogeneity planning, we asked for
position papers.
We got over 100 of them. Parsing them has been a real challenge for the
organizers. We should think about what approach we might employ for
gathering input from the community prior to the workshop and what we
would do with that material prior to the workshop
vs. incorporating into the final report.
Workshop. We will shoot for around 60 attendees, maybe 30/20/10
lab/university/industry. This helps keep the justification
and planning fairly simple. We would plan for a 1.75-day workshop to
facilitate covering five topical areas as separate, single-track
discussions. If for some reason we end up needing only 1.5 days for the
meeting proper, this would leave more time for the
organizers to meet and discuss report completion.
We will (mostly) assume that attendees have read the FSD prior to arrival
at the meeting. Never the less, we will spend 1/2 day bringing the
attendees up to speed, focusing on the background material from the
FSD. This is similar to the opening plenary session portion of the BES
workshop structure.
The remainder of the meeting could follow the model from the 2014 SSIO
meeting: 90-minute moderated discussion sessions, with notes captured
live in PPT. With 1.75 days we can accommodate 5 total discussion
topics. We will place a topic of broad interest last to encourage
attendees to participate for the full meeting. These sessions will
identify current status in the area, its CS challenges, its
opportunities, and its research needs, reflecting the breakout
session portion of the BES workshop structure. Because the meeting is
single track, there is no need for outbrief sessions (i.e., the
participants will have been present for all discussions).
Report. The final report will be generated by augmenting the FSD with
material summarizing the outcomes of the workshop.
We will target a 75-100 page document (the 2015 report was approximately
90 pages). The outline for the report will look something like this:
- Executive Summary
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps (need to understand what this typically includes)
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- CS challenges and opportunities
- Priority research directions (3-5)
- Topic 2 ...
- Priority research directions
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
<2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-doc-20180124.pdf
1
0
It might be nice to make sure we cover/count the storage facilities. I
actually think it will be covered without much challenge, but someone like
Dave Dillow might be a nice add in that he used to work for a DOE
facility, and now works for a large cloud facility.
Just a thought.
Cheers,
Brad
On 4/9/18, 5:32 PM, "ssio-2018 on behalf of Mohror, Kathryn"
<ssio-2018-bounces(a)lists.mcs.anl.gov on behalf of mohror1(a)llnl.gov> wrote:
>Hi Scott,
>
>I think I/O fits under ‘Storage Systems’.
>
>Kathryn
>
>From: Klasky, Scott A. <klasky(a)ornl.gov>
>Sent: Monday, April 9, 2018 4:27 PM
>To: Mohror, Kathryn <mohror1(a)llnl.gov>; Ross, Robert B.
><rross(a)mcs.anl.gov>; ssio-2018(a)lists.mcs.anl.gov
>Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
>
>
>
>Why don't we have an io column? I vote to have this column
>
>
>
>
>
>
>
>Scott
>
>
>
>
>
>
>
>
>________________________________________
>
>
>From: Mohror, Kathryn <mohror1(a)llnl.gov>
>Date: April 9, 2018 at 11:54:47 PM GMT+2
>To: Ross, Robert B. <rross(a)mcs.anl.gov>,
>ssio-2018(a)lists.mcs.anl.gov <ssio-2018(a)lists.mcs.anl.gov>
>Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
>
>
>Hi everybody,
>
>I checked the spreadsheet of potential attendees and no one has entered
>any yet!
>
>
>Rob, are we including only DOE folks? Or are university and industry
>people included as well?
>
>Please take a bit of time to start to jot down some names.
>The spreadsheet is here:
>https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxzC
>gRjPj9YEY/edit?usp=sharing
><https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxz
>CgRjPj9YEY/edit?usp=sharing>
>
>Kathryn
>
>From: ssio-2018 <ssio-2018-bounces(a)lists.mcs.anl.gov>
>On Behalf Of Ross, Robert B.
>Sent: Monday, April 9, 2018 11:33 AM
>To: ssio-2018(a)lists.mcs.anl.gov
>Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
>
>
>
>Hi all,
>
>
>Thanks for taking the time to fill out the polls. While we don't have the
>go-ahead for the virtual summer or workshop yet, let's hold the following
>dates:
>
>- May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans,
>contact info TBD)
>
>- August 1-2 for a tentative workshop (probably in DC/MD)
>
>
>
>Thanks!
>
>
>
>Rob
>
>
>On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
>
>
>Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for
>getting one out of two?). Please fill out the Doodle polls when you can.
>-- Rob
>
>
>On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
>
>
>Hi all,
>
>
>
>Scott asked a good question that I will summarize as "what is going on?".
>Here's what I know:
>
>
>
>Virtual Summit. Lucy has asked for permission to have our virtual summit,
>back on the 21st. I will double-check with her on status of that request.
>Meanwhile we should
> all gather our thoughts for the summit:
>
>- Think about what you think the big issues are in SSIO research for
>enabling DOE science, talk with your team and colleagues.
>
>- Fill out this poll; we're optimistic that we will have permission to do
>this soon:
>
>
>https://doodle.com/poll/mvfxb8fcqucimtim
><https://doodle.com/poll/mvfxb8fcqucimtim>
>
>
>
>Scope. We've been asked to think about the data and learning applications
>in addition to traditional simulation. I don't think this is a surprise,
>but it does mean
> that we need to think about those communities as we think about who to
>include in the discussion/workshop and what to have in our initial report.
>
>
>
>Organizers. Gary has asked to take a step back but is happy to chime in
>as needed. Here's our revised team:
>
>- Brad Settlemyer (LANL)
>
>- Scott Klasky (ORNL)
>
>- Glenn Lockwood (LBNL)
>
>- Kathryn Mohror (LLNL)
>
>- Lee Ward (SNL)
>
>
>
>I think everyone probably knows Brad. I'm sure he'll do a great job. Gary
>will remain on the mailing list, too.
>
>
>
>Meeting Dates. Lucy and I have chatted on possible dates, and we want to
>get some dates on the calendar. We're looking for two days in
>July/August. Here's a short
> set of dates that work for me and Lucy; we're hoping to get lucky and
>get a hit. Note that we didn't include Mon. or Fri. in these so people
>can travel during the week:
>
>-
>https://doodle.com/poll/nw4xsqpx2agd3dhh
><https://doodle.com/poll/nw4xsqpx2agd3dhh>
>
>
>
>Attendees. We should start thinking about attendees. I've attached the
>list of people from the SSIO workshop back in late 2014 as a starting
>point. Can someone volunteer
> to create a Google spreadsheet that we could use for organizing possible
>participants. Note that we don't have permission to have the meeting at
>this point, so we're really just discussing who we think key thinkers are
>in the area.
>
>
>
>I'm also attaching the EH and ML pre-workshop documents as examples of
>the kind of report we need to prepare.
>
>
>
>Thanks,
>
>
>
>Rob
>
>---
>
>
>
>
>On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV> wrote:
>
>
>Rob tasked me with trying to kick this off. I had something different in
>mind but I suspect a call is a better idea. Thanks!
>Sent from my iPhone
>
>
>On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov> wrote:
>
>
>Hi Rob,
>
>Has there been a doodle poll for this?
>
>
>
>Scott
>
>
>
>From: ssio-2018 [mailto:[email protected]] On
> Behalf Of Ross, Robert B.
>Sent: Friday, January 12, 2018 1:18 PM
>To: ssio-2018(a)lists.mcs.anl.gov
>Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
>
>
>
>
>
>Forgot the attachment, of course. -- Rob
>
>
>
>
>
>
>
>
>
>On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov> wrote:
>
>
>
>
>Hi all,
>
>
>
>
>
>I've chatted with all of you previously about this, but I've been lax in
>getting moving. Lucy has asked me to organize a workshop to help define
>priority research directions for scalable storage and I/O (SSIO),
> ideally leading to a call for proposals some time in 2018. Additionally,
>there is some interest from Almadena on the NSF side to restart an
>inter-agency activity in this space. We will be the organizers that drive
>the ASCR process forward. We'll see also how
> things develop on the inter-agency front and what role we might play in
>that process.
>
>
>
>
>
>Our goal is to organize our thoughts about future research directions,
>produce an initial document capturing that thinking and describing state
>of the art, plan and hold a workshop to identify a set of priority
> research directions, and produce a final workshop report capturing this
>information.
>
>
>
>
>
>My initial plan is based off the experiences from the old HEC-FSIO
>process, the more recent 2014 ASCR SSIO workshop (for which there is a
>2015 report), and the most recent Extreme Heterogeneity workshop planning
> (for which we will have the workshop soon). All of this is up for
>discussion/negotiation.
>
>
>
>
>
>Summary and action items.
>
>
>- Think about what you think the big issues are in SSIO research for
>enabling DOE science, talk with your team and colleagues.
>
>
>- Start putting together the "virtual" summit materials.
>
>
>- Look for the Doodle poll for a summit date.
>
>
>- Look for a Doodle poll for a time we could regularly meet to discuss
>progress.
>
>
>
>
>
>Details follow. Thanks in advance for your help in making this a success!
>
>
>
>
>
>Rob
>
>
>---
>
>
>
>
>
>Organizers. Here's our team:
>
>
>- Gary Grider (LANL)
>
>
>- Scott Klasky (ORNL)
>
>
>- Glenn Lockwood (LBNL)
>
>
>- Kathryn Mohror (LLNL)
>
>
>- Lee Ward (SNL)
>
>
>
>
>
>Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be
>included and possibly take over as contact before we are done.
>
>
>
>
>
>We have a mailing list: ssio-2018(a)lists.mcs.anl.gov for
> discussion. You have been added to the list. Please let me know ASAP if
>for some reason you cannot help.
>
>
>
>
>
>
>
>
>Topical areas and initial discussion. Five focus areas were discussed for
>further investigation at the prior workshop.
>
>
>- HW/SW architectures
>
>
>- Metadata, name spaces, and provenance
>
>
>- Supporting science data
>
>
>- Integrating with external services
>
>
>- Understanding SSIO systems
>
>
>
>
>
>I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th
>of a day, rather than flying somewhere for a day to meet in person. Each
>organizer would present their view on challenges as relate
> to mission and discuss if/how they think we should adjust the topical
>areas in response to changes since 2015. I think of these as 20 minute
>presentations that are devoid of our own research specifics/propaganda
>and focus on these big picture items. From this
> we would select a set of topical areas used to organize the factual
>status document (see below).
>
>
>
>
>
>
>
>
>Factual status document. We are modeling our process after the BES "basic
>research needs" approach (see attachment), with
> adaptations to the scope of the area and ASCR needs. This process
>includes the generation of a factual status document (FSD) prior to the
>workshop. There is a separate group working down this path for an Extreme
>Heterogeneity FSD and workshop, too. The FSD
> can be thought of as the background sections of the final workshop
>document, and that's how we will approach it.
>
>
>
>
>
>With this in mind, we will organize the document around our initial
>guesses at topical areas, understanding that the topical areas may have
>to be adjusted in the final report as a result of discussion. The
> goal would be to complete the following for the FSD:
>
>
>
>
>
>- Introduction
>
>
> - Science Challenges
>
>
> - Technology Challenges
>
>
> - Knowledge Gaps -- what we need to know that we don't know
>
>
>- Research Needs
>
>
> - Topic 1
>
>
> - Definition/summary of area
>
>
> - Current status
>
>
> - Topic 2 ...
>
>
>- Appendix: References
>
>
>
>
>
>This is likely to be a 50 page document.
>
>
>
>
>
>This document would be partitioned into sections to be managed by
>individual organizers. Rather than requesting white papers from potential
>attendees, organizers would invite experts in respective areas to
> participate in the development of requisite sections. It might be
>easiest to develop this document using an online system such as Google
>Docs, understanding that eventually it would be turned into a Word
>document and finally into a PDF for publication as the
> workshop report.
>
>
>
>
>
>
>
>
>Position papers. In the Extreme Heterogeneity planning, we asked for
>position papers.
> We got over 100 of them. Parsing them has been a real challenge for the
>organizers. We should think about what approach we might employ for
>gathering input from the community prior to the workshop and what we
>would do with that material prior to the workshop
> vs. incorporating into the final report.
>
>
>
>
>
>
>
>
>Workshop. We will shoot for around 60 attendees, maybe 30/20/10
>lab/university/industry. This helps keep the justification
> and planning fairly simple. We would plan for a 1.75-day workshop to
>facilitate covering five topical areas as separate, single-track
>discussions. If for some reason we end up needing only 1.5 days for the
>meeting proper, this would leave more time for the
> organizers to meet and discuss report completion.
>
>
>
>
>
>We will (mostly) assume that attendees have read the FSD prior to arrival
>at the meeting. Never the less, we will spend 1/2 day bringing the
>attendees up to speed, focusing on the background material from the
> FSD. This is similar to the opening plenary session portion of the BES
>workshop structure.
>
>
>
>
>
>The remainder of the meeting could follow the model from the 2014 SSIO
>meeting: 90-minute moderated discussion sessions, with notes captured
>live in PPT. With 1.75 days we can accommodate 5 total discussion
> topics. We will place a topic of broad interest last to encourage
>attendees to participate for the full meeting. These sessions will
>identify current status in the area, its CS challenges, its
>opportunities, and its research needs, reflecting the breakout
> session portion of the BES workshop structure. Because the meeting is
>single track, there is no need for outbrief sessions (i.e., the
>participants will have been present for all discussions).
>
>
>
>
>
>
>
>
>Report. The final report will be generated by augmenting the FSD with
>material summarizing the outcomes of the workshop.
> We will target a 75-100 page document (the 2015 report was approximately
>90 pages). The outline for the report will look something like this:
>
>
>
>
>
>- Executive Summary
>
>
>- Introduction
>
>
> - Science Challenges
>
>
> - Technology Challenges
>
>
> - Knowledge Gaps (need to understand what this typically includes)
>
>
>- Research Needs
>
>
> - Topic 1
>
>
> - Definition/summary of area
>
>
> - Current status
>
>
> - CS challenges and opportunities
>
>
> - Priority research directions (3-5)
>
>
> - Topic 2 ...
>
>
>- Priority research directions
>
>
>
>
>
>
>_______________________________________________
>ssio-2018 mailing list
>ssio-2018(a)lists.mcs.anl.gov
>https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
>
>
>
>
>
>
>
>
>
>
>
>
>
>_______________________________________________
>ssio-2018 mailing list
>ssio-2018(a)lists.mcs.anl.gov
>https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
>
>
>
>
>
>
>
>
>
>
><2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-doc-20180124.pdf
>>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
1
0
Hi Scott,
I think I/O fits under 'Storage Systems'.
Kathryn
From: Klasky, Scott A. <klasky(a)ornl.gov>
Sent: Monday, April 9, 2018 4:27 PM
To: Mohror, Kathryn <mohror1(a)llnl.gov>; Ross, Robert B. <rross(a)mcs.anl.gov>; ssio-2018(a)lists.mcs.anl.gov
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Why don't we have an io column? I vote to have this column
Scott
________________________________
From: Mohror, Kathryn <mohror1(a)llnl.gov<mailto:[email protected]>>
Date: April 9, 2018 at 11:54:47 PM GMT+2
To: Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>>, ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> <ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi everybody,
I checked the spreadsheet of potential attendees and no one has entered any yet!
Rob, are we including only DOE folks? Or are university and industry people included as well?
Please take a bit of time to start to jot down some names.
The spreadsheet is here: https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxzCg…
Kathryn
From: ssio-2018 <ssio-2018-bounces(a)lists.mcs.anl.gov<mailto:[email protected]>> On Behalf Of Ross, Robert B.
Sent: Monday, April 9, 2018 11:33 AM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi all,
Thanks for taking the time to fill out the polls. While we don't have the go-ahead for the virtual summer or workshop yet, let's hold the following dates:
- May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans, contact info TBD)
- August 1-2 for a tentative workshop (probably in DC/MD)
Thanks!
Rob
On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for getting one out of two?). Please fill out the Doodle polls when you can. -- Rob
On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
Scott asked a good question that I will summarize as "what is going on?". Here's what I know:
Virtual Summit. Lucy has asked for permission to have our virtual summit, back on the 21st. I will double-check with her on status of that request. Meanwhile we should all gather our thoughts for the summit:
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Fill out this poll; we're optimistic that we will have permission to do this soon:
https://doodle.com/poll/mvfxb8fcqucimtim
Scope. We've been asked to think about the data and learning applications in addition to traditional simulation. I don't think this is a surprise, but it does mean that we need to think about those communities as we think about who to include in the discussion/workshop and what to have in our initial report.
Organizers. Gary has asked to take a step back but is happy to chime in as needed. Here's our revised team:
- Brad Settlemyer (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
I think everyone probably knows Brad. I'm sure he'll do a great job. Gary will remain on the mailing list, too.
Meeting Dates. Lucy and I have chatted on possible dates, and we want to get some dates on the calendar. We're looking for two days in July/August. Here's a short set of dates that work for me and Lucy; we're hoping to get lucky and get a hit. Note that we didn't include Mon. or Fri. in these so people can travel during the week:
- https://doodle.com/poll/nw4xsqpx2agd3dhh
Attendees. We should start thinking about attendees. I've attached the list of people from the SSIO workshop back in late 2014 as a starting point. Can someone volunteer to create a Google spreadsheet that we could use for organizing possible participants. Note that we don't have permission to have the meeting at this point, so we're really just discussing who we think key thinkers are in the area.
I'm also attaching the EH and ML pre-workshop documents as examples of the kind of report we need to prepare.
Thanks,
Rob
---
On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV<mailto:[email protected]>> wrote:
Rob tasked me with trying to kick this off. I had something different in mind but I suspect a call is a better idea. Thanks!
Sent from my iPhone
On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov<mailto:[email protected]>> wrote:
Hi Rob,
Has there been a doodle poll for this?
Scott
From: ssio-2018 [mailto:[email protected]] On Behalf Of Ross, Robert B.
Sent: Friday, January 12, 2018 1:18 PM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Forgot the attachment, of course. -- Rob
On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
I've chatted with all of you previously about this, but I've been lax in getting moving. Lucy has asked me to organize a workshop to help define priority research directions for scalable storage and I/O (SSIO), ideally leading to a call for proposals some time in 2018. Additionally, there is some interest from Almadena on the NSF side to restart an inter-agency activity in this space. We will be the organizers that drive the ASCR process forward. We'll see also how things develop on the inter-agency front and what role we might play in that process.
Our goal is to organize our thoughts about future research directions, produce an initial document capturing that thinking and describing state of the art, plan and hold a workshop to identify a set of priority research directions, and produce a final workshop report capturing this information.
My initial plan is based off the experiences from the old HEC-FSIO process, the more recent 2014 ASCR SSIO workshop (for which there is a 2015 report), and the most recent Extreme Heterogeneity workshop planning (for which we will have the workshop soon). All of this is up for discussion/negotiation.
Summary and action items.
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Start putting together the "virtual" summit materials.
- Look for the Doodle poll for a summit date.
- Look for a Doodle poll for a time we could regularly meet to discuss progress.
Details follow. Thanks in advance for your help in making this a success!
Rob
---
Organizers. Here's our team:
- Gary Grider (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be included and possibly take over as contact before we are done.
We have a mailing list: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> for discussion. You have been added to the list. Please let me know ASAP if for some reason you cannot help.
Topical areas and initial discussion. Five focus areas were discussed for further investigation at the prior workshop.
- HW/SW architectures
- Metadata, name spaces, and provenance
- Supporting science data
- Integrating with external services
- Understanding SSIO systems
I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th of a day, rather than flying somewhere for a day to meet in person. Each organizer would present their view on challenges as relate to mission and discuss if/how they think we should adjust the topical areas in response to changes since 2015. I think of these as 20 minute presentations that are devoid of our own research specifics/propaganda and focus on these big picture items. From this we would select a set of topical areas used to organize the factual status document (see below).
Factual status document. We are modeling our process after the BES "basic research needs" approach (see attachment), with adaptations to the scope of the area and ASCR needs. This process includes the generation of a factual status document (FSD) prior to the workshop. There is a separate group working down this path for an Extreme Heterogeneity FSD and workshop, too. The FSD can be thought of as the background sections of the final workshop document, and that's how we will approach it.
With this in mind, we will organize the document around our initial guesses at topical areas, understanding that the topical areas may have to be adjusted in the final report as a result of discussion. The goal would be to complete the following for the FSD:
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps -- what we need to know that we don't know
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- Topic 2 ...
- Appendix: References
This is likely to be a 50 page document.
This document would be partitioned into sections to be managed by individual organizers. Rather than requesting white papers from potential attendees, organizers would invite experts in respective areas to participate in the development of requisite sections. It might be easiest to develop this document using an online system such as Google Docs, understanding that eventually it would be turned into a Word document and finally into a PDF for publication as the workshop report.
Position papers. In the Extreme Heterogeneity planning, we asked for position papers. We got over 100 of them. Parsing them has been a real challenge for the organizers. We should think about what approach we might employ for gathering input from the community prior to the workshop and what we would do with that material prior to the workshop vs. incorporating into the final report.
Workshop. We will shoot for around 60 attendees, maybe 30/20/10 lab/university/industry. This helps keep the justification and planning fairly simple. We would plan for a 1.75-day workshop to facilitate covering five topical areas as separate, single-track discussions. If for some reason we end up needing only 1.5 days for the meeting proper, this would leave more time for the organizers to meet and discuss report completion.
We will (mostly) assume that attendees have read the FSD prior to arrival at the meeting. Never the less, we will spend 1/2 day bringing the attendees up to speed, focusing on the background material from the FSD. This is similar to the opening plenary session portion of the BES workshop structure.
The remainder of the meeting could follow the model from the 2014 SSIO meeting: 90-minute moderated discussion sessions, with notes captured live in PPT. With 1.75 days we can accommodate 5 total discussion topics. We will place a topic of broad interest last to encourage attendees to participate for the full meeting. These sessions will identify current status in the area, its CS challenges, its opportunities, and its research needs, reflecting the breakout session portion of the BES workshop structure. Because the meeting is single track, there is no need for outbrief sessions (i.e., the participants will have been present for all discussions).
Report. The final report will be generated by augmenting the FSD with material summarizing the outcomes of the workshop. We will target a 75-100 page document (the 2015 report was approximately 90 pages). The outline for the report will look something like this:
- Executive Summary
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps (need to understand what this typically includes)
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- CS challenges and opportunities
- Priority research directions (3-5)
- Topic 2 ...
- Priority research directions
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
<2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-doc-20180124.pdf>
1
0
Why don't we have an io column? I vote to have this column
Scott
________________________________
From: Mohror, Kathryn <mohror1(a)llnl.gov>
Date: April 9, 2018 at 11:54:47 PM GMT+2
To: Ross, Robert B. <rross(a)mcs.anl.gov>, ssio-2018(a)lists.mcs.anl.gov <ssio-2018(a)lists.mcs.anl.gov>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi everybody,
I checked the spreadsheet of potential attendees and no one has entered any yet!
Rob, are we including only DOE folks? Or are university and industry people included as well?
Please take a bit of time to start to jot down some names.
The spreadsheet is here: https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxzCg…
Kathryn
From: ssio-2018 <ssio-2018-bounces(a)lists.mcs.anl.gov> On Behalf Of Ross, Robert B.
Sent: Monday, April 9, 2018 11:33 AM
To: ssio-2018(a)lists.mcs.anl.gov
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi all,
Thanks for taking the time to fill out the polls. While we don't have the go-ahead for the virtual summer or workshop yet, let's hold the following dates:
- May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans, contact info TBD)
- August 1-2 for a tentative workshop (probably in DC/MD)
Thanks!
Rob
On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for getting one out of two?). Please fill out the Doodle polls when you can. -- Rob
On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
Scott asked a good question that I will summarize as "what is going on?". Here's what I know:
Virtual Summit. Lucy has asked for permission to have our virtual summit, back on the 21st. I will double-check with her on status of that request. Meanwhile we should all gather our thoughts for the summit:
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Fill out this poll; we're optimistic that we will have permission to do this soon:
https://doodle.com/poll/mvfxb8fcqucimtim
Scope. We've been asked to think about the data and learning applications in addition to traditional simulation. I don't think this is a surprise, but it does mean that we need to think about those communities as we think about who to include in the discussion/workshop and what to have in our initial report.
Organizers. Gary has asked to take a step back but is happy to chime in as needed. Here's our revised team:
- Brad Settlemyer (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
I think everyone probably knows Brad. I'm sure he'll do a great job. Gary will remain on the mailing list, too.
Meeting Dates. Lucy and I have chatted on possible dates, and we want to get some dates on the calendar. We're looking for two days in July/August. Here's a short set of dates that work for me and Lucy; we're hoping to get lucky and get a hit. Note that we didn't include Mon. or Fri. in these so people can travel during the week:
- https://doodle.com/poll/nw4xsqpx2agd3dhh
Attendees. We should start thinking about attendees. I've attached the list of people from the SSIO workshop back in late 2014 as a starting point. Can someone volunteer to create a Google spreadsheet that we could use for organizing possible participants. Note that we don't have permission to have the meeting at this point, so we're really just discussing who we think key thinkers are in the area.
I'm also attaching the EH and ML pre-workshop documents as examples of the kind of report we need to prepare.
Thanks,
Rob
---
On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV<mailto:[email protected]>> wrote:
Rob tasked me with trying to kick this off. I had something different in mind but I suspect a call is a better idea. Thanks!
Sent from my iPhone
On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov<mailto:[email protected]>> wrote:
Hi Rob,
Has there been a doodle poll for this?
Scott
From: ssio-2018 [mailto:[email protected]] On Behalf Of Ross, Robert B.
Sent: Friday, January 12, 2018 1:18 PM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Forgot the attachment, of course. -- Rob
On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
I've chatted with all of you previously about this, but I've been lax in getting moving. Lucy has asked me to organize a workshop to help define priority research directions for scalable storage and I/O (SSIO), ideally leading to a call for proposals some time in 2018. Additionally, there is some interest from Almadena on the NSF side to restart an inter-agency activity in this space. We will be the organizers that drive the ASCR process forward. We'll see also how things develop on the inter-agency front and what role we might play in that process.
Our goal is to organize our thoughts about future research directions, produce an initial document capturing that thinking and describing state of the art, plan and hold a workshop to identify a set of priority research directions, and produce a final workshop report capturing this information.
My initial plan is based off the experiences from the old HEC-FSIO process, the more recent 2014 ASCR SSIO workshop (for which there is a 2015 report), and the most recent Extreme Heterogeneity workshop planning (for which we will have the workshop soon). All of this is up for discussion/negotiation.
Summary and action items.
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Start putting together the "virtual" summit materials.
- Look for the Doodle poll for a summit date.
- Look for a Doodle poll for a time we could regularly meet to discuss progress.
Details follow. Thanks in advance for your help in making this a success!
Rob
---
Organizers. Here's our team:
- Gary Grider (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be included and possibly take over as contact before we are done.
We have a mailing list: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> for discussion. You have been added to the list. Please let me know ASAP if for some reason you cannot help.
Topical areas and initial discussion. Five focus areas were discussed for further investigation at the prior workshop.
- HW/SW architectures
- Metadata, name spaces, and provenance
- Supporting science data
- Integrating with external services
- Understanding SSIO systems
I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th of a day, rather than flying somewhere for a day to meet in person. Each organizer would present their view on challenges as relate to mission and discuss if/how they think we should adjust the topical areas in response to changes since 2015. I think of these as 20 minute presentations that are devoid of our own research specifics/propaganda and focus on these big picture items. From this we would select a set of topical areas used to organize the factual status document (see below).
Factual status document. We are modeling our process after the BES "basic research needs" approach (see attachment), with adaptations to the scope of the area and ASCR needs. This process includes the generation of a factual status document (FSD) prior to the workshop. There is a separate group working down this path for an Extreme Heterogeneity FSD and workshop, too. The FSD can be thought of as the background sections of the final workshop document, and that's how we will approach it.
With this in mind, we will organize the document around our initial guesses at topical areas, understanding that the topical areas may have to be adjusted in the final report as a result of discussion. The goal would be to complete the following for the FSD:
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps -- what we need to know that we don't know
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- Topic 2 ...
- Appendix: References
This is likely to be a 50 page document.
This document would be partitioned into sections to be managed by individual organizers. Rather than requesting white papers from potential attendees, organizers would invite experts in respective areas to participate in the development of requisite sections. It might be easiest to develop this document using an online system such as Google Docs, understanding that eventually it would be turned into a Word document and finally into a PDF for publication as the workshop report.
Position papers. In the Extreme Heterogeneity planning, we asked for position papers. We got over 100 of them. Parsing them has been a real challenge for the organizers. We should think about what approach we might employ for gathering input from the community prior to the workshop and what we would do with that material prior to the workshop vs. incorporating into the final report.
Workshop. We will shoot for around 60 attendees, maybe 30/20/10 lab/university/industry. This helps keep the justification and planning fairly simple. We would plan for a 1.75-day workshop to facilitate covering five topical areas as separate, single-track discussions. If for some reason we end up needing only 1.5 days for the meeting proper, this would leave more time for the organizers to meet and discuss report completion.
We will (mostly) assume that attendees have read the FSD prior to arrival at the meeting. Never the less, we will spend 1/2 day bringing the attendees up to speed, focusing on the background material from the FSD. This is similar to the opening plenary session portion of the BES workshop structure.
The remainder of the meeting could follow the model from the 2014 SSIO meeting: 90-minute moderated discussion sessions, with notes captured live in PPT. With 1.75 days we can accommodate 5 total discussion topics. We will place a topic of broad interest last to encourage attendees to participate for the full meeting. These sessions will identify current status in the area, its CS challenges, its opportunities, and its research needs, reflecting the breakout session portion of the BES workshop structure. Because the meeting is single track, there is no need for outbrief sessions (i.e., the participants will have been present for all discussions).
Report. The final report will be generated by augmenting the FSD with material summarizing the outcomes of the workshop. We will target a 75-100 page document (the 2015 report was approximately 90 pages). The outline for the report will look something like this:
- Executive Summary
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps (need to understand what this typically includes)
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- CS challenges and opportunities
- Priority research directions (3-5)
- Topic 2 ...
- Priority research directions
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
<2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-doc-20180124.pdf>
1
0
Hi everybody,
I checked the spreadsheet of potential attendees and no one has entered any yet!
Rob, are we including only DOE folks? Or are university and industry people included as well?
Please take a bit of time to start to jot down some names.
The spreadsheet is here: https://docs.google.com/spreadsheets/d/16oeV2PAaOdeuKzjWJ1M8tF45mgzSSMuxzCg…
Kathryn
From: ssio-2018 <ssio-2018-bounces(a)lists.mcs.anl.gov> On Behalf Of Ross, Robert B.
Sent: Monday, April 9, 2018 11:33 AM
To: ssio-2018(a)lists.mcs.anl.gov
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Hi all,
Thanks for taking the time to fill out the polls. While we don't have the go-ahead for the virtual summer or workshop yet, let's hold the following dates:
- May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans, contact info TBD)
- August 1-2 for a tentative workshop (probably in DC/MD)
Thanks!
Rob
On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for getting one out of two?). Please fill out the Doodle polls when you can. -- Rob
On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
Scott asked a good question that I will summarize as "what is going on?". Here's what I know:
Virtual Summit. Lucy has asked for permission to have our virtual summit, back on the 21st. I will double-check with her on status of that request. Meanwhile we should all gather our thoughts for the summit:
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Fill out this poll; we're optimistic that we will have permission to do this soon:
https://doodle.com/poll/mvfxb8fcqucimtim
Scope. We've been asked to think about the data and learning applications in addition to traditional simulation. I don't think this is a surprise, but it does mean that we need to think about those communities as we think about who to include in the discussion/workshop and what to have in our initial report.
Organizers. Gary has asked to take a step back but is happy to chime in as needed. Here's our revised team:
- Brad Settlemyer (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
I think everyone probably knows Brad. I'm sure he'll do a great job. Gary will remain on the mailing list, too.
Meeting Dates. Lucy and I have chatted on possible dates, and we want to get some dates on the calendar. We're looking for two days in July/August. Here's a short set of dates that work for me and Lucy; we're hoping to get lucky and get a hit. Note that we didn't include Mon. or Fri. in these so people can travel during the week:
- https://doodle.com/poll/nw4xsqpx2agd3dhh
Attendees. We should start thinking about attendees. I've attached the list of people from the SSIO workshop back in late 2014 as a starting point. Can someone volunteer to create a Google spreadsheet that we could use for organizing possible participants. Note that we don't have permission to have the meeting at this point, so we're really just discussing who we think key thinkers are in the area.
I'm also attaching the EH and ML pre-workshop documents as examples of the kind of report we need to prepare.
Thanks,
Rob
---
On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV<mailto:[email protected]>> wrote:
Rob tasked me with trying to kick this off. I had something different in mind but I suspect a call is a better idea. Thanks!
Sent from my iPhone
On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov<mailto:[email protected]>> wrote:
Hi Rob,
Has there been a doodle poll for this?
Scott
From: ssio-2018 [mailto:[email protected]] On Behalf Of Ross, Robert B.
Sent: Friday, January 12, 2018 1:18 PM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Forgot the attachment, of course. -- Rob
On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
I've chatted with all of you previously about this, but I've been lax in getting moving. Lucy has asked me to organize a workshop to help define priority research directions for scalable storage and I/O (SSIO), ideally leading to a call for proposals some time in 2018. Additionally, there is some interest from Almadena on the NSF side to restart an inter-agency activity in this space. We will be the organizers that drive the ASCR process forward. We'll see also how things develop on the inter-agency front and what role we might play in that process.
Our goal is to organize our thoughts about future research directions, produce an initial document capturing that thinking and describing state of the art, plan and hold a workshop to identify a set of priority research directions, and produce a final workshop report capturing this information.
My initial plan is based off the experiences from the old HEC-FSIO process, the more recent 2014 ASCR SSIO workshop (for which there is a 2015 report), and the most recent Extreme Heterogeneity workshop planning (for which we will have the workshop soon). All of this is up for discussion/negotiation.
Summary and action items.
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Start putting together the "virtual" summit materials.
- Look for the Doodle poll for a summit date.
- Look for a Doodle poll for a time we could regularly meet to discuss progress.
Details follow. Thanks in advance for your help in making this a success!
Rob
---
Organizers. Here's our team:
- Gary Grider (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be included and possibly take over as contact before we are done.
We have a mailing list: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> for discussion. You have been added to the list. Please let me know ASAP if for some reason you cannot help.
Topical areas and initial discussion. Five focus areas were discussed for further investigation at the prior workshop.
- HW/SW architectures
- Metadata, name spaces, and provenance
- Supporting science data
- Integrating with external services
- Understanding SSIO systems
I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th of a day, rather than flying somewhere for a day to meet in person. Each organizer would present their view on challenges as relate to mission and discuss if/how they think we should adjust the topical areas in response to changes since 2015. I think of these as 20 minute presentations that are devoid of our own research specifics/propaganda and focus on these big picture items. From this we would select a set of topical areas used to organize the factual status document (see below).
Factual status document. We are modeling our process after the BES "basic research needs" approach (see attachment), with adaptations to the scope of the area and ASCR needs. This process includes the generation of a factual status document (FSD) prior to the workshop. There is a separate group working down this path for an Extreme Heterogeneity FSD and workshop, too. The FSD can be thought of as the background sections of the final workshop document, and that's how we will approach it.
With this in mind, we will organize the document around our initial guesses at topical areas, understanding that the topical areas may have to be adjusted in the final report as a result of discussion. The goal would be to complete the following for the FSD:
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps -- what we need to know that we don't know
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- Topic 2 ...
- Appendix: References
This is likely to be a 50 page document.
This document would be partitioned into sections to be managed by individual organizers. Rather than requesting white papers from potential attendees, organizers would invite experts in respective areas to participate in the development of requisite sections. It might be easiest to develop this document using an online system such as Google Docs, understanding that eventually it would be turned into a Word document and finally into a PDF for publication as the workshop report.
Position papers. In the Extreme Heterogeneity planning, we asked for position papers. We got over 100 of them. Parsing them has been a real challenge for the organizers. We should think about what approach we might employ for gathering input from the community prior to the workshop and what we would do with that material prior to the workshop vs. incorporating into the final report.
Workshop. We will shoot for around 60 attendees, maybe 30/20/10 lab/university/industry. This helps keep the justification and planning fairly simple. We would plan for a 1.75-day workshop to facilitate covering five topical areas as separate, single-track discussions. If for some reason we end up needing only 1.5 days for the meeting proper, this would leave more time for the organizers to meet and discuss report completion.
We will (mostly) assume that attendees have read the FSD prior to arrival at the meeting. Never the less, we will spend 1/2 day bringing the attendees up to speed, focusing on the background material from the FSD. This is similar to the opening plenary session portion of the BES workshop structure.
The remainder of the meeting could follow the model from the 2014 SSIO meeting: 90-minute moderated discussion sessions, with notes captured live in PPT. With 1.75 days we can accommodate 5 total discussion topics. We will place a topic of broad interest last to encourage attendees to participate for the full meeting. These sessions will identify current status in the area, its CS challenges, its opportunities, and its research needs, reflecting the breakout session portion of the BES workshop structure. Because the meeting is single track, there is no need for outbrief sessions (i.e., the participants will have been present for all discussions).
Report. The final report will be generated by augmenting the FSD with material summarizing the outcomes of the workshop. We will target a 75-100 page document (the 2015 report was approximately 90 pages). The outline for the report will look something like this:
- Executive Summary
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps (need to understand what this typically includes)
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- CS challenges and opportunities
- Priority research directions (3-5)
- Topic 2 ...
- Priority research directions
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
<2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-doc-20180124.pdf>
1
0
Hi all,
Thanks for taking the time to fill out the polls. While we don't have the go-ahead for the virtual summer or workshop yet, let's hold the following dates:
- May 9, 10:00am - 4:30pm central for a virtual summit (Blue Jeans, contact info TBD)
- August 1-2 for a tentative workshop (probably in DC/MD)
Thanks!
Rob
On Mar 12, 2018, at 11:38 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Thanks to Gary and Kathryn for filling out the Doodle polls (and Brad for getting one out of two?). Please fill out the Doodle polls when you can. -- Rob
On Mar 8, 2018, at 8:02 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
Scott asked a good question that I will summarize as "what is going on?". Here's what I know:
Virtual Summit. Lucy has asked for permission to have our virtual summit, back on the 21st. I will double-check with her on status of that request. Meanwhile we should all gather our thoughts for the summit:
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Fill out this poll; we're optimistic that we will have permission to do this soon:
https://doodle.com/poll/mvfxb8fcqucimtim
Scope. We've been asked to think about the data and learning applications in addition to traditional simulation. I don't think this is a surprise, but it does mean that we need to think about those communities as we think about who to include in the discussion/workshop and what to have in our initial report.
Organizers. Gary has asked to take a step back but is happy to chime in as needed. Here's our revised team:
- Brad Settlemyer (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
I think everyone probably knows Brad. I'm sure he'll do a great job. Gary will remain on the mailing list, too.
Meeting Dates. Lucy and I have chatted on possible dates, and we want to get some dates on the calendar. We're looking for two days in July/August. Here's a short set of dates that work for me and Lucy; we're hoping to get lucky and get a hit. Note that we didn't include Mon. or Fri. in these so people can travel during the week:
- https://doodle.com/poll/nw4xsqpx2agd3dhh
Attendees. We should start thinking about attendees. I've attached the list of people from the SSIO workshop back in late 2014 as a starting point. Can someone volunteer to create a Google spreadsheet that we could use for organizing possible participants. Note that we don't have permission to have the meeting at this point, so we're really just discussing who we think key thinkers are in the area.
I'm also attaching the EH and ML pre-workshop documents as examples of the kind of report we need to prepare.
Thanks,
Rob
---
On Mar 1, 2018, at 4:17 PM, Ward, Lee <lee(a)SANDIA.GOV<mailto:[email protected]>> wrote:
Rob tasked me with trying to kick this off. I had something different in mind but I suspect a call is a better idea. Thanks!
Sent from my iPhone
On Mar 1, 2018, at 2:34 PM, Klasky, Scott A. <klasky(a)ornl.gov<mailto:[email protected]>> wrote:
Hi Rob,
Has there been a doodle poll for this?
Scott
From: ssio-2018 [mailto:[email protected]] On Behalf Of Ross, Robert B.
Sent: Friday, January 12, 2018 1:18 PM
To: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
Subject: Re: [Ssio-2018] SSIO 2018 Workshop Planning
Forgot the attachment, of course. -- Rob
On Jan 11, 2018, at 10:32 AM, Ross, Robert B. <rross(a)mcs.anl.gov<mailto:[email protected]>> wrote:
Hi all,
I've chatted with all of you previously about this, but I've been lax in getting moving. Lucy has asked me to organize a workshop to help define priority research directions for scalable storage and I/O (SSIO), ideally leading to a call for proposals some time in 2018. Additionally, there is some interest from Almadena on the NSF side to restart an inter-agency activity in this space. We will be the organizers that drive the ASCR process forward. We'll see also how things develop on the inter-agency front and what role we might play in that process.
Our goal is to organize our thoughts about future research directions, produce an initial document capturing that thinking and describing state of the art, plan and hold a workshop to identify a set of priority research directions, and produce a final workshop report capturing this information.
My initial plan is based off the experiences from the old HEC-FSIO process, the more recent 2014 ASCR SSIO workshop (for which there is a 2015 report), and the most recent Extreme Heterogeneity workshop planning (for which we will have the workshop soon). All of this is up for discussion/negotiation.
Summary and action items.
- Think about what you think the big issues are in SSIO research for enabling DOE science, talk with your team and colleagues.
- Start putting together the "virtual" summit materials.
- Look for the Doodle poll for a summit date.
- Look for a Doodle poll for a time we could regularly meet to discuss progress.
Details follow. Thanks in advance for your help in making this a success!
Rob
---
Organizers. Here's our team:
- Gary Grider (LANL)
- Scott Klasky (ORNL)
- Glenn Lockwood (LBNL)
- Kathryn Mohror (LLNL)
- Lee Ward (SNL)
Lucy Nowell is our ASCR contact, and I expect that Laura Biven will be included and possibly take over as contact before we are done.
We have a mailing list: ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]> for discussion. You have been added to the list. Please let me know ASAP if for some reason you cannot help.
Topical areas and initial discussion. Five focus areas were discussed for further investigation at the prior workshop.
- HW/SW architectures
- Metadata, name spaces, and provenance
- Supporting science data
- Integrating with external services
- Understanding SSIO systems
I propose that we do a "virtual" summit (i.e., videoconference) for 3/4th of a day, rather than flying somewhere for a day to meet in person. Each organizer would present their view on challenges as relate to mission and discuss if/how they think we should adjust the topical areas in response to changes since 2015. I think of these as 20 minute presentations that are devoid of our own research specifics/propaganda and focus on these big picture items. From this we would select a set of topical areas used to organize the factual status document (see below).
Factual status document. We are modeling our process after the BES "basic research needs" approach (see attachment), with adaptations to the scope of the area and ASCR needs. This process includes the generation of a factual status document (FSD) prior to the workshop. There is a separate group working down this path for an Extreme Heterogeneity FSD and workshop, too. The FSD can be thought of as the background sections of the final workshop document, and that's how we will approach it.
With this in mind, we will organize the document around our initial guesses at topical areas, understanding that the topical areas may have to be adjusted in the final report as a result of discussion. The goal would be to complete the following for the FSD:
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps -- what we need to know that we don't know
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- Topic 2 ...
- Appendix: References
This is likely to be a 50 page document.
This document would be partitioned into sections to be managed by individual organizers. Rather than requesting white papers from potential attendees, organizers would invite experts in respective areas to participate in the development of requisite sections. It might be easiest to develop this document using an online system such as Google Docs, understanding that eventually it would be turned into a Word document and finally into a PDF for publication as the workshop report.
Position papers. In the Extreme Heterogeneity planning, we asked for position papers. We got over 100 of them. Parsing them has been a real challenge for the organizers. We should think about what approach we might employ for gathering input from the community prior to the workshop and what we would do with that material prior to the workshop vs. incorporating into the final report.
Workshop. We will shoot for around 60 attendees, maybe 30/20/10 lab/university/industry. This helps keep the justification and planning fairly simple. We would plan for a 1.75-day workshop to facilitate covering five topical areas as separate, single-track discussions. If for some reason we end up needing only 1.5 days for the meeting proper, this would leave more time for the organizers to meet and discuss report completion.
We will (mostly) assume that attendees have read the FSD prior to arrival at the meeting. Never the less, we will spend 1/2 day bringing the attendees up to speed, focusing on the background material from the FSD. This is similar to the opening plenary session portion of the BES workshop structure.
The remainder of the meeting could follow the model from the 2014 SSIO meeting: 90-minute moderated discussion sessions, with notes captured live in PPT. With 1.75 days we can accommodate 5 total discussion topics. We will place a topic of broad interest last to encourage attendees to participate for the full meeting. These sessions will identify current status in the area, its CS challenges, its opportunities, and its research needs, reflecting the breakout session portion of the BES workshop structure. Because the meeting is single track, there is no need for outbrief sessions (i.e., the participants will have been present for all discussions).
Report. The final report will be generated by augmenting the FSD with material summarizing the outcomes of the workshop. We will target a 75-100 page document (the 2015 report was approximately 90 pages). The outline for the report will look something like this:
- Executive Summary
- Introduction
- Science Challenges
- Technology Challenges
- Knowledge Gaps (need to understand what this typically includes)
- Research Needs
- Topic 1
- Definition/summary of area
- Current status
- CS challenges and opportunities
- Priority research directions (3-5)
- Topic 2 ...
- Priority research directions
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
_______________________________________________
ssio-2018 mailing list
ssio-2018(a)lists.mcs.anl.gov<mailto:[email protected]>
https://lists.mcs.anl.gov/mailman/listinfo/ssio-2018
<2018-01-15-eh-pre-workshop-document.pdf><ml-pre-workshop-doc-20180124.pdf>
1
0