Difference between revisions of "EPSCI Group Meeting Mar. 22, 2021"
Jump to navigation
Jump to search
Line 131: | Line 131: | ||
** Oracle is dropping support for some packages causing alternatives to be explored | ** Oracle is dropping support for some packages causing alternatives to be explored | ||
*** Vardan considering moving to web-based and possible hand-held. Wants to discuss with others first as it will take some effort to implement. | *** Vardan considering moving to web-based and possible hand-held. Wants to discuss with others first as it will take some effort to implement. | ||
+ | |||
+ | * AI | ||
+ | ** GPU purchase finalized | ||
+ | *** 3 nodes of 16x Nvidia T4 cards each | ||
+ | **** Last may be half full due to budget constraint | ||
+ | ** Hydra paper | ||
+ | *** Multiple comment sets generated. | ||
+ | ** Jupyterhub | ||
+ | *** Some issues with custom kernels | ||
+ | ** AI Experimental Controls | ||
+ | *** Thomas working on roadmap for project work | ||
+ | *** Diana starts May 1st | ||
+ | **** Preliminary capstone project sketch submitted | ||
+ | |||
+ | * OSG | ||
+ | ** Issue identified | ||
+ | *** Log files for condor being written to /volatile disk | ||
+ | *** Culprits notified and rectifying situation | ||
+ | ** scosg20 will not have lustre mounted at all to eliminate these types of issues | ||
+ | ** working to pull lustre from scosg16 as well. |
Latest revision as of 13:51, 29 March 2021
The meeting time is 10:00am.
Connection Info:
You can connect using BlueJeans Video conferencing (ID: 253 300 597). (Click "Expand" to the right for details -->):
Meeting URL https://bluejeans.com/253300597?src=join_info Meeting ID 253 300 597 Want to dial in from a phone? Dial one of the following numbers: +1.888.240.2560 (US Toll Free) (see all numbers - https://www.bluejeans.com/premium-numbers) Enter the meeting ID and passcode followed by # Connecting from a room system? Dial: bjn.vc or 199.48.152.152 and enter your meeting ID & passcode
Agenda:
- Previous meeting
- Announcements
- DE-FOA-0002490 $50k-$2M for 2 years. Meeting Thur@2pm bluejeans
- SC_FOA_0002482 $100k-$2M/year for 3 years. Meeting Mon@3pm
- Fortnight paper for Mar. 29: HEP-Frame: Improving the efficiency of pipelined data transformation & filtering for scientific analyses
- Conferences and Workshops
- Workshop: CFNS-ANL Joint Workshop on Instrumenting the 2nd IR at the EIC (Mar. 17-19)
- SEA'S IMPROVING SCIENTIFIC SOFTWARE CONFERENCE AND TUTORIALS 2021 (Mar. 22-26)
- Vardan: Streaming data processing from multiple satellite data sets under the NASA/GEWEX SRB project. (Mar. 26 @ 1pm)
- Autonomous Discovery in Science and Engineering workshop (April 20-22)
- vCHEP2021 (May 17-21)
- Thomas, Kishan: Hydra
- Vardan, Nathan, David (+Hall-B, Fast Electronics, and TriDAS groups): TriDAS + JANA2 SRO
- David: HOSS!
- ACAT2021 (Nov. 29 - Dec. 3)
- Ongoing Activities
- Scientific Software support
- JLab Common Environment (CE) + SPACK
- ServiceNow (mapmanager, fputil, fpack, bos, bankdef)
- HOMEWORK ASSIGNMENT: Everyone please read the Quickstart Instructions and test the system. Report any issues.
- EIC
- ACTS
- Collaboration with ANL
- Offline frameworks (CLARA, JANA2)
- JLab Common Environment (CE) + SPACK
- Data Transport
- Meetings Tue., Wed., and Fri.
- DAQ systems
- SRO
- SAMPA + ERSAP + JANA2 + INDRA-ASTRA = April 1st
- CODA (CODA3 support, EVIO-6)
- SRO
- A.I.
- GPU purchase for ENP
- Hydra paper
- Jupyterhub + GPU
- Experimental Controls
- Offsite Computing
- NERSC, PSC, IU
- XSEDE application for PSC bridges-2 declined
- OSG
- NERSC, PSC, IU
- Scientific Software support
- AOT
Minutes:
Attendees: David L., Carl T., Nathan B., Kishan R., Vardan G., Thomas B., Mike G.
- Announcements
- SPACK
- CODA packages basically ready to go
- Need to set permissions on epsci-spack repository on GitHub to allow others to commit
- Some issue with last few packages from JLabCE 2.4
- build system not fully contained in source tarballs
- ACTS
- Upgraded to 6.2
- Nathan is going through the new tutorials which give a lot of info. Going through them in order to make sure he doesn't missing anything.
- Running some simple test cases provided by Dmitry. Not all of them working yet.
- Able to identify small bug in JANA2 making the exercise useful to JANA2 development and not just EIC/ACTS
- Close to point where genfit/RAVE can be completely abandoned
- Upgraded to 6.2
- JANA2
- David put in a pull request for some Python API support changes and some minor changes to core classes.
- CLARA
- clara1602 and clara1603 are two of three computers that were purchased with a grant Vardan had to work on the NASA project a few years back
- Computers were in the farm, but not being utilized. Bryan pulled them and repurposed them as dedicated ifarm machines for CLAS
- CLAS12 production processing on farm is underway (or very soon to be if not already)
- clara1602 and clara1603 are two of three computers that were purchased with a grant Vardan had to work on the NASA project a few years back
- Data Transport
- Met with fast electronics group and ESnet folks last week
- Need to purchase pair of Xilinx U280 cards (ESnet people said cards we had were too old)
- They were interested in broader area use cases (e.g. EIC to JLab)
- More discussions are needed on JLab side to converge on some clearer specs.
- Mike has scheduled a few meetings with various players this week to try and work this out
- Mike looked up that Dark fiber refers to unused fibers in a bundle that have already been strung.
- It looks like there is some excess capacity out there we may be able to use
- Eventually will get copy of firmware from the ESnet guys we will use for initial testing.
- SRO
- Vardan continues VTP system performance studies
- Plans to introduce Data Lake to try and "cool down" the stream to eliminate all frame loss at higher rates.
- CODA
- Oracle is dropping support for some packages causing alternatives to be explored
- Vardan considering moving to web-based and possible hand-held. Wants to discuss with others first as it will take some effort to implement.
- Oracle is dropping support for some packages causing alternatives to be explored
- AI
- GPU purchase finalized
- 3 nodes of 16x Nvidia T4 cards each
- Last may be half full due to budget constraint
- 3 nodes of 16x Nvidia T4 cards each
- Hydra paper
- Multiple comment sets generated.
- Jupyterhub
- Some issues with custom kernels
- AI Experimental Controls
- Thomas working on roadmap for project work
- Diana starts May 1st
- Preliminary capstone project sketch submitted
- GPU purchase finalized
- OSG
- Issue identified
- Log files for condor being written to /volatile disk
- Culprits notified and rectifying situation
- scosg20 will not have lustre mounted at all to eliminate these types of issues
- working to pull lustre from scosg16 as well.
- Issue identified