Hauptseite > Publikationsdatenbank > Massively parallel loading > print

001     141704
005     20250317091728.0
020 _ _ |a 978-1-4503-2130-3
024 7 _ |a 10.1145/2464996.2465020
|2 doi
024 7 _ |a altmetric:21822711
|2 altmetric
037 _ _ |a FZJ-2014-00071
041 _ _ |a eng
100 1 _ |a Frings, Wolfgang
|0 P:(DE-Juel1)132108
|b 0
|u fzj
|e Corresponding author
111 2 _ |a the 27th international ACM conference
|c Eugene
|d 2013-06-10 - 2013-06-14
|w Oregon
245 _ _ |a Massively parallel loading
260 _ _ |a New York, NY
|c 2013
|b ACM Press New York, New York, USA
295 1 0 |a Proceedings of the 27th international ACM conference on International conference on supercomputing - ICS '13
300 _ _ |a 389-398
336 7 _ |a Contribution to a conference proceedings
|b contrib
|m contrib
|0 PUB:(DE-HGF)8
|s 1389192531_18637
|2 PUB:(DE-HGF)
336 7 _ |a Contribution to a book
|0 PUB:(DE-HGF)7
|2 PUB:(DE-HGF)
|m contb
336 7 _ |a Conference Paper
|0 33
|2 EndNote
336 7 _ |a CONFERENCE_PAPER
|2 ORCID
336 7 _ |a Output Types/Conference Paper
|2 DataCite
336 7 _ |a conferenceObject
|2 DRIVER
336 7 _ |a INPROCEEDINGS
|2 BibTeX
520 _ _ |a Dynamic linking has many advantages for managing large code bases, but dynamically linked applications have not typically scaled well on high performance computing systems. Splitting a monolithic executable into many dynamic shared object (DSO) files decreases compile time for large codes, reduces runtime memory requirements by allowing modules to be loaded and unloaded as needed, and allows common DSOs to be shared among many executables. However, launching an executable that depends on many DSOs causes a flood of file system operations at program start-up, when each process in the parallel application loads its dependencies. At large scales, this operation has an effect similar to a site-wide denial-of-service attack, as even large parallel file systems struggle to service so many simultaneous requests. In this paper, we present SPINDLE, a novel approach to parallel loading that coordinates simultaneous file system operations with a scalable network of cache server processes. Our approach is transparent to user applications. We extend the GNU loader, which is used in Linux as well as proprietary operating systems, to limit the number of simultaneous file system operations, quickly loading DSOs without thrashing the file system. Our experiments show that our prototype implementation has a low overhead and increases the scalability of Pynamic, a benchmark that stresses the dynamic loader, by a factor of 20.
536 _ _ |a 41G - Supercomputer Facility (POF2-41G21)
|0 G:(DE-HGF)POF2-41G21
|c POF2-41G21
|x 0
|f POF II
536 _ _ |0 G:(DE-Juel-1)ATMLAO
|a ATMLAO - ATML Application Optimization and User Service Tools (ATMLAO)
|c ATMLAO
|x 1
588 _ _ |a Dataset connected to CrossRef Conference
700 1 _ |a Ahn, Dong H.
|0 P:(DE-HGF)0
|b 1
700 1 _ |a LeGendre, Matthew
|0 P:(DE-HGF)0
|b 2
700 1 _ |a Gamblin, Todd
|0 P:(DE-HGF)0
|b 3
700 1 _ |a de Supinski, Bronis R.
|0 P:(DE-HGF)0
|b 4
700 1 _ |a Wolf, Felix
|0 P:(DE-Juel1)132299
|b 5
|u fzj
773 _ _ |a 10.1145/2464996.2465020
909 C O |o oai:juser.fz-juelich.de:141704
|p VDB
910 1 _ |a Forschungszentrum Jülich GmbH
|0 I:(DE-588b)5008462-8
|k FZJ
|b 0
|6 P:(DE-Juel1)132108
910 1 _ |a Forschungszentrum Jülich GmbH
|0 I:(DE-588b)5008462-8
|k FZJ
|b 5
|6 P:(DE-Juel1)132299
913 2 _ |a DE-HGF
|b Key Technologies
|l Supercomputing & Big Data
|1 G:(DE-HGF)POF3-510
|0 G:(DE-HGF)POF3-513
|2 G:(DE-HGF)POF3-500
|v Supercomputer Facility
|x 0
913 1 _ |a DE-HGF
|b Schlüsseltechnologien
|l Supercomputing
|1 G:(DE-HGF)POF2-410
|0 G:(DE-HGF)POF2-41G21
|2 G:(DE-HGF)POF2-400
|v Supercomputer Facility
|x 0
|4 G:(DE-HGF)POF
|3 G:(DE-HGF)POF2
914 1 _ |y 2013
920 _ _ |l yes
920 1 _ |0 I:(DE-Juel1)JSC-20090406
|k JSC
|l Jülich Supercomputing Center
|x 0
980 _ _ |a contrib
980 _ _ |a VDB
980 _ _ |a UNRESTRICTED
980 _ _ |a contb
980 _ _ |a I:(DE-Juel1)JSC-20090406

LibraryCollectionCLSMajorCLSMinorLanguageAuthor
Marc 21