Network slicing to improve multicasting in HPC clusters
Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The u...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Alsmadi, Izzat [verfasserIn] |
|---|
| Format: |
Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2018 |
|---|
| Schlagwörter: |
|---|
| Anmerkung: |
© Springer Science+Business Media, LLC, part of Springer Nature 2018 |
|---|
| Übergeordnetes Werk: |
Enthalten in: Cluster computing - Springer US, 1998, 21(2018), 3 vom: 31. Jan., Seite 1493-1506 |
|---|---|
| Übergeordnetes Werk: |
volume:21 ; year:2018 ; number:3 ; day:31 ; month:01 ; pages:1493-1506 |
| Links: |
|---|
| DOI / URN: |
10.1007/s10586-017-1561-5 |
|---|
| Katalog-ID: |
OLC2066390852 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | OLC2066390852 | ||
| 003 | DE-627 | ||
| 005 | 20230503024840.0 | ||
| 007 | tu | ||
| 008 | 200819s2018 xx ||||| 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s10586-017-1561-5 |2 doi | |
| 035 | |a (DE-627)OLC2066390852 | ||
| 035 | |a (DE-He213)s10586-017-1561-5-p | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 004 |q VZ |
| 084 | |a 54.50$jProgrammierung: Allgemeines |2 bkl | ||
| 084 | |a 54.32$jRechnerkommunikation |2 bkl | ||
| 084 | |a 54.25$jParallele Datenverarbeitung |2 bkl | ||
| 100 | 1 | |a Alsmadi, Izzat |e verfasserin |0 (orcid)0000-0001-7832-5081 |4 aut | |
| 245 | 1 | 0 | |a Network slicing to improve multicasting in HPC clusters |
| 264 | 1 | |c 2018 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ohne Hilfsmittel zu benutzen |b n |2 rdamedia | ||
| 338 | |a Band |b nc |2 rdacarrier | ||
| 500 | |a © Springer Science+Business Media, LLC, part of Springer Nature 2018 | ||
| 520 | |a Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. | ||
| 650 | 4 | |a HPC | |
| 650 | 4 | |a Network slicing | |
| 650 | 4 | |a SDN | |
| 700 | 1 | |a Khreishah, Abdallah |4 aut | |
| 700 | 1 | |a Xu, Dianxiang |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Cluster computing |d Springer US, 1998 |g 21(2018), 3 vom: 31. Jan., Seite 1493-1506 |w (DE-627)265187907 |w (DE-600)1465290-0 |w (DE-576)9265187905 |x 1386-7857 |7 nnns |
| 773 | 1 | 8 | |g volume:21 |g year:2018 |g number:3 |g day:31 |g month:01 |g pages:1493-1506 |
| 856 | 4 | 1 | |u https://doi.org/10.1007/s10586-017-1561-5 |z lizenzpflichtig |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a SYSFLAG_A | ||
| 912 | |a GBV_OLC | ||
| 912 | |a SSG-OLC-MAT | ||
| 912 | |a GBV_ILN_70 | ||
| 936 | b | k | |a 54.50$jProgrammierung: Allgemeines |q VZ |0 181569876 |0 (DE-625)181569876 |
| 936 | b | k | |a 54.32$jRechnerkommunikation |q VZ |0 10640623X |0 (DE-625)10640623X |
| 936 | b | k | |a 54.25$jParallele Datenverarbeitung |q VZ |0 181569892 |0 (DE-625)181569892 |
| 951 | |a AR | ||
| 952 | |d 21 |j 2018 |e 3 |b 31 |c 01 |h 1493-1506 | ||
| author_variant |
i a ia a k ak d x dx |
|---|---|
| matchkey_str |
article:13867857:2018----::ewrsiigomrvmliatni |
| hierarchy_sort_str |
2018 |
| bklnumber |
54.50$jProgrammierung: Allgemeines 54.32$jRechnerkommunikation 54.25$jParallele Datenverarbeitung |
| publishDate |
2018 |
| allfields |
10.1007/s10586-017-1561-5 doi (DE-627)OLC2066390852 (DE-He213)s10586-017-1561-5-p DE-627 ger DE-627 rakwb eng 004 VZ 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung bkl Alsmadi, Izzat verfasserin (orcid)0000-0001-7832-5081 aut Network slicing to improve multicasting in HPC clusters 2018 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC, part of Springer Nature 2018 Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. HPC Network slicing SDN Khreishah, Abdallah aut Xu, Dianxiang aut Enthalten in Cluster computing Springer US, 1998 21(2018), 3 vom: 31. Jan., Seite 1493-1506 (DE-627)265187907 (DE-600)1465290-0 (DE-576)9265187905 1386-7857 nnns volume:21 year:2018 number:3 day:31 month:01 pages:1493-1506 https://doi.org/10.1007/s10586-017-1561-5 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT GBV_ILN_70 54.50$jProgrammierung: Allgemeines VZ 181569876 (DE-625)181569876 54.32$jRechnerkommunikation VZ 10640623X (DE-625)10640623X 54.25$jParallele Datenverarbeitung VZ 181569892 (DE-625)181569892 AR 21 2018 3 31 01 1493-1506 |
| spelling |
10.1007/s10586-017-1561-5 doi (DE-627)OLC2066390852 (DE-He213)s10586-017-1561-5-p DE-627 ger DE-627 rakwb eng 004 VZ 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung bkl Alsmadi, Izzat verfasserin (orcid)0000-0001-7832-5081 aut Network slicing to improve multicasting in HPC clusters 2018 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC, part of Springer Nature 2018 Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. HPC Network slicing SDN Khreishah, Abdallah aut Xu, Dianxiang aut Enthalten in Cluster computing Springer US, 1998 21(2018), 3 vom: 31. Jan., Seite 1493-1506 (DE-627)265187907 (DE-600)1465290-0 (DE-576)9265187905 1386-7857 nnns volume:21 year:2018 number:3 day:31 month:01 pages:1493-1506 https://doi.org/10.1007/s10586-017-1561-5 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT GBV_ILN_70 54.50$jProgrammierung: Allgemeines VZ 181569876 (DE-625)181569876 54.32$jRechnerkommunikation VZ 10640623X (DE-625)10640623X 54.25$jParallele Datenverarbeitung VZ 181569892 (DE-625)181569892 AR 21 2018 3 31 01 1493-1506 |
| allfields_unstemmed |
10.1007/s10586-017-1561-5 doi (DE-627)OLC2066390852 (DE-He213)s10586-017-1561-5-p DE-627 ger DE-627 rakwb eng 004 VZ 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung bkl Alsmadi, Izzat verfasserin (orcid)0000-0001-7832-5081 aut Network slicing to improve multicasting in HPC clusters 2018 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC, part of Springer Nature 2018 Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. HPC Network slicing SDN Khreishah, Abdallah aut Xu, Dianxiang aut Enthalten in Cluster computing Springer US, 1998 21(2018), 3 vom: 31. Jan., Seite 1493-1506 (DE-627)265187907 (DE-600)1465290-0 (DE-576)9265187905 1386-7857 nnns volume:21 year:2018 number:3 day:31 month:01 pages:1493-1506 https://doi.org/10.1007/s10586-017-1561-5 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT GBV_ILN_70 54.50$jProgrammierung: Allgemeines VZ 181569876 (DE-625)181569876 54.32$jRechnerkommunikation VZ 10640623X (DE-625)10640623X 54.25$jParallele Datenverarbeitung VZ 181569892 (DE-625)181569892 AR 21 2018 3 31 01 1493-1506 |
| allfieldsGer |
10.1007/s10586-017-1561-5 doi (DE-627)OLC2066390852 (DE-He213)s10586-017-1561-5-p DE-627 ger DE-627 rakwb eng 004 VZ 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung bkl Alsmadi, Izzat verfasserin (orcid)0000-0001-7832-5081 aut Network slicing to improve multicasting in HPC clusters 2018 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC, part of Springer Nature 2018 Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. HPC Network slicing SDN Khreishah, Abdallah aut Xu, Dianxiang aut Enthalten in Cluster computing Springer US, 1998 21(2018), 3 vom: 31. Jan., Seite 1493-1506 (DE-627)265187907 (DE-600)1465290-0 (DE-576)9265187905 1386-7857 nnns volume:21 year:2018 number:3 day:31 month:01 pages:1493-1506 https://doi.org/10.1007/s10586-017-1561-5 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT GBV_ILN_70 54.50$jProgrammierung: Allgemeines VZ 181569876 (DE-625)181569876 54.32$jRechnerkommunikation VZ 10640623X (DE-625)10640623X 54.25$jParallele Datenverarbeitung VZ 181569892 (DE-625)181569892 AR 21 2018 3 31 01 1493-1506 |
| allfieldsSound |
10.1007/s10586-017-1561-5 doi (DE-627)OLC2066390852 (DE-He213)s10586-017-1561-5-p DE-627 ger DE-627 rakwb eng 004 VZ 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung bkl Alsmadi, Izzat verfasserin (orcid)0000-0001-7832-5081 aut Network slicing to improve multicasting in HPC clusters 2018 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer Science+Business Media, LLC, part of Springer Nature 2018 Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. HPC Network slicing SDN Khreishah, Abdallah aut Xu, Dianxiang aut Enthalten in Cluster computing Springer US, 1998 21(2018), 3 vom: 31. Jan., Seite 1493-1506 (DE-627)265187907 (DE-600)1465290-0 (DE-576)9265187905 1386-7857 nnns volume:21 year:2018 number:3 day:31 month:01 pages:1493-1506 https://doi.org/10.1007/s10586-017-1561-5 lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT GBV_ILN_70 54.50$jProgrammierung: Allgemeines VZ 181569876 (DE-625)181569876 54.32$jRechnerkommunikation VZ 10640623X (DE-625)10640623X 54.25$jParallele Datenverarbeitung VZ 181569892 (DE-625)181569892 AR 21 2018 3 31 01 1493-1506 |
| language |
English |
| source |
Enthalten in Cluster computing 21(2018), 3 vom: 31. Jan., Seite 1493-1506 volume:21 year:2018 number:3 day:31 month:01 pages:1493-1506 |
| sourceStr |
Enthalten in Cluster computing 21(2018), 3 vom: 31. Jan., Seite 1493-1506 volume:21 year:2018 number:3 day:31 month:01 pages:1493-1506 |
| format_phy_str_mv |
Article |
| institution |
findex.gbv.de |
| topic_facet |
HPC Network slicing SDN |
| dewey-raw |
004 |
| isfreeaccess_bool |
false |
| container_title |
Cluster computing |
| authorswithroles_txt_mv |
Alsmadi, Izzat @@aut@@ Khreishah, Abdallah @@aut@@ Xu, Dianxiang @@aut@@ |
| publishDateDaySort_date |
2018-01-31T00:00:00Z |
| hierarchy_top_id |
265187907 |
| dewey-sort |
14 |
| id |
OLC2066390852 |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">OLC2066390852</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230503024840.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">200819s2018 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s10586-017-1561-5</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2066390852</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s10586-017-1561-5-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">004</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.50$jProgrammierung: Allgemeines</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.32$jRechnerkommunikation</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.25$jParallele Datenverarbeitung</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Alsmadi, Izzat</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0001-7832-5081</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Network slicing to improve multicasting in HPC clusters</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2018</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© Springer Science+Business Media, LLC, part of Springer Nature 2018</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">HPC</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Network slicing</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">SDN</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Khreishah, Abdallah</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Xu, Dianxiang</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Cluster computing</subfield><subfield code="d">Springer US, 1998</subfield><subfield code="g">21(2018), 3 vom: 31. Jan., Seite 1493-1506</subfield><subfield code="w">(DE-627)265187907</subfield><subfield code="w">(DE-600)1465290-0</subfield><subfield code="w">(DE-576)9265187905</subfield><subfield code="x">1386-7857</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:21</subfield><subfield code="g">year:2018</subfield><subfield code="g">number:3</subfield><subfield code="g">day:31</subfield><subfield code="g">month:01</subfield><subfield code="g">pages:1493-1506</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s10586-017-1561-5</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.50$jProgrammierung: Allgemeines</subfield><subfield code="q">VZ</subfield><subfield code="0">181569876</subfield><subfield code="0">(DE-625)181569876</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.32$jRechnerkommunikation</subfield><subfield code="q">VZ</subfield><subfield code="0">10640623X</subfield><subfield code="0">(DE-625)10640623X</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.25$jParallele Datenverarbeitung</subfield><subfield code="q">VZ</subfield><subfield code="0">181569892</subfield><subfield code="0">(DE-625)181569892</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">21</subfield><subfield code="j">2018</subfield><subfield code="e">3</subfield><subfield code="b">31</subfield><subfield code="c">01</subfield><subfield code="h">1493-1506</subfield></datafield></record></collection>
|
| author |
Alsmadi, Izzat |
| spellingShingle |
Alsmadi, Izzat ddc 004 bkl 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung misc HPC misc Network slicing misc SDN Network slicing to improve multicasting in HPC clusters |
| authorStr |
Alsmadi, Izzat |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)265187907 |
| format |
Article |
| dewey-ones |
004 - Data processing & computer science |
| delete_txt_mv |
keep |
| author_role |
aut aut aut |
| collection |
OLC |
| remote_str |
false |
| illustrated |
Not Illustrated |
| issn |
1386-7857 |
| topic_title |
004 VZ 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung bkl Network slicing to improve multicasting in HPC clusters HPC Network slicing SDN |
| topic |
ddc 004 bkl 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung misc HPC misc Network slicing misc SDN |
| topic_unstemmed |
ddc 004 bkl 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung misc HPC misc Network slicing misc SDN |
| topic_browse |
ddc 004 bkl 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung misc HPC misc Network slicing misc SDN |
| format_facet |
Aufsätze Gedruckte Aufsätze |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
nc |
| hierarchy_parent_title |
Cluster computing |
| hierarchy_parent_id |
265187907 |
| dewey-tens |
000 - Computer science, knowledge & systems |
| hierarchy_top_title |
Cluster computing |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)265187907 (DE-600)1465290-0 (DE-576)9265187905 |
| title |
Network slicing to improve multicasting in HPC clusters |
| ctrlnum |
(DE-627)OLC2066390852 (DE-He213)s10586-017-1561-5-p |
| title_full |
Network slicing to improve multicasting in HPC clusters |
| author_sort |
Alsmadi, Izzat |
| journal |
Cluster computing |
| journalStr |
Cluster computing |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
000 - Computer science, information & general works |
| recordtype |
marc |
| publishDateSort |
2018 |
| contenttype_str_mv |
txt |
| container_start_page |
1493 |
| author_browse |
Alsmadi, Izzat Khreishah, Abdallah Xu, Dianxiang |
| container_volume |
21 |
| class |
004 VZ 54.50$jProgrammierung: Allgemeines bkl 54.32$jRechnerkommunikation bkl 54.25$jParallele Datenverarbeitung bkl |
| format_se |
Aufsätze |
| author-letter |
Alsmadi, Izzat |
| doi_str_mv |
10.1007/s10586-017-1561-5 |
| normlink |
(ORCID)0000-0001-7832-5081 181569876 10640623X 181569892 |
| normlink_prefix_str_mv |
(orcid)0000-0001-7832-5081 181569876 (DE-625)181569876 10640623X (DE-625)10640623X 181569892 (DE-625)181569892 |
| dewey-full |
004 |
| title_sort |
network slicing to improve multicasting in hpc clusters |
| title_auth |
Network slicing to improve multicasting in HPC clusters |
| abstract |
Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. © Springer Science+Business Media, LLC, part of Springer Nature 2018 |
| abstractGer |
Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. © Springer Science+Business Media, LLC, part of Springer Nature 2018 |
| abstract_unstemmed |
Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters. © Springer Science+Business Media, LLC, part of Springer Nature 2018 |
| collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-MAT GBV_ILN_70 |
| container_issue |
3 |
| title_short |
Network slicing to improve multicasting in HPC clusters |
| url |
https://doi.org/10.1007/s10586-017-1561-5 |
| remote_bool |
false |
| author2 |
Khreishah, Abdallah Xu, Dianxiang |
| author2Str |
Khreishah, Abdallah Xu, Dianxiang |
| ppnlink |
265187907 |
| mediatype_str_mv |
n |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| doi_str |
10.1007/s10586-017-1561-5 |
| up_date |
2024-07-04T04:25:41.402Z |
| _version_ |
1803621122861694976 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">OLC2066390852</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230503024840.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">200819s2018 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s10586-017-1561-5</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2066390852</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s10586-017-1561-5-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">004</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.50$jProgrammierung: Allgemeines</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.32$jRechnerkommunikation</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.25$jParallele Datenverarbeitung</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Alsmadi, Izzat</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0001-7832-5081</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Network slicing to improve multicasting in HPC clusters</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2018</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© Springer Science+Business Media, LLC, part of Springer Nature 2018</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract In high performance computing (HPC) resources’ extensive experiments are frequently executed. HPC resources (e.g. computing machines and switches) should be able to handle running several experiments in parallel. Typically HPC utilizes parallelization in programs, processing and data. The underlying network is seen as the only non-parallelized HPC component (i.e. no dynamic virtual slicing based on HPC jobs). In this scope we present an approach in this paper to utilize software defined networking (SDN) to parallelize HPC clusters among the different running experiments. We propose to accomplish this through two major components: A passive module (network mapper/remapper) to select for each experiment as soon as it starts the least busy resources in the network, and an SDN-HPC active load balancer to perform more complex and intelligent operations. Active load balancer can logically divide the network based on experiments’ host files. The goal is to reduce traffic to unnecessary hosts or ports. An HPC experiment should multicast, rather than broadcast to only cluster nodes that are used by the experiment. We use virtual tenant network modules in Opendaylight controller to create VLANs based on HPC experiments. In each HPC host, virtual interfaces are created to isolate traffic from the different experiments. The traffic between the different physical hosts that belong to the same experiment can be distinguished based on the VLAN ID assigned to each experiment. We evaluate the new approach using several HPC public benchmarks. Results show a significant enhancement in experiments’ performance especially when HPC cluster experiences running several heavy load experiments simultaneously. Results show also that this multi-casting approach can significantly reduce casting overhead that is caused by using a single cast for all resources in the HPC cluster. In comparison with InfiniBand networks that offer interconnect services with low latency and high bandwidth, HPC services based on SDN can provide two distinguished objectives that may not be possible with InfiniBand: The first objective is the integration of HPC with Ethernet enterprise networks and hence expanding HPC usage to much wider domains. The second objective is the ability to enable users and their applications to customize HPC services with different QoS requirements that fit the different needs of those applications and optimize the usage of HPC clusters.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">HPC</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Network slicing</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">SDN</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Khreishah, Abdallah</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Xu, Dianxiang</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Cluster computing</subfield><subfield code="d">Springer US, 1998</subfield><subfield code="g">21(2018), 3 vom: 31. Jan., Seite 1493-1506</subfield><subfield code="w">(DE-627)265187907</subfield><subfield code="w">(DE-600)1465290-0</subfield><subfield code="w">(DE-576)9265187905</subfield><subfield code="x">1386-7857</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:21</subfield><subfield code="g">year:2018</subfield><subfield code="g">number:3</subfield><subfield code="g">day:31</subfield><subfield code="g">month:01</subfield><subfield code="g">pages:1493-1506</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s10586-017-1561-5</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.50$jProgrammierung: Allgemeines</subfield><subfield code="q">VZ</subfield><subfield code="0">181569876</subfield><subfield code="0">(DE-625)181569876</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.32$jRechnerkommunikation</subfield><subfield code="q">VZ</subfield><subfield code="0">10640623X</subfield><subfield code="0">(DE-625)10640623X</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.25$jParallele Datenverarbeitung</subfield><subfield code="q">VZ</subfield><subfield code="0">181569892</subfield><subfield code="0">(DE-625)181569892</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">21</subfield><subfield code="j">2018</subfield><subfield code="e">3</subfield><subfield code="b">31</subfield><subfield code="c">01</subfield><subfield code="h">1493-1506</subfield></datafield></record></collection>
|
| score |
7.399584 |