SlideShare a Scribd company logo
1 of 51
Download to read offline
Loom:FlexibleandEfficientNICPacket
Scheduling
ReadingCircle
M2TakuyaFukuoka
November27,2020
1
Aboutthispaper
PublishedinNSDI'19.
PresentLoom,anewNICdesignthatmovesallper-flow
schedulingdecisionsoutoftheOSandintotheNIC
Sourcecodeandofficialslideisavailable
Iborrowedmanyfiguresandsentencesfromtheofficialslide
inthispresentation
2
Datacentersandcloud
Datacentershavemanycompetingtenants,whichshouldbe
isolated
WithVMsandcontainers,itiscurrentlypossibletoensurethat
tenantsfairlyshareCPUandmemory
Ontheotherhand,networkisolationisproblematicespecially
wheneachtenantrunavarietyofapplicationswithdifferent
performanceneeds
Latency-sensitiveapplicationsuchasmemcached
Throughput-sensitiveapplicationssuchasHadoopandSpark
3
Thegoal
Networkoperatorsmustspecifyandenforceanetworkisolation
policy,whichrequiresapporopriatepacketsscheduling
4
Coremotivationofthiswork
TheOSisresponsiblefordynamicallyenforcingthesepolicies
However,itcannotfullyachievethegoalbecausetoday'sNICis
notsoflexible
Proposeloom,whichmovesallper-flowschedulingdecisionsout
oftheOSandintotheNIC
5
Outlines
Motivation
Loom'sDesign
.Anewnetworkpolicyabstraction:restricteddirectedacyclic
graphs(DAGs)
.Aprogrammablehierarchicalpacketscheduler
.AnewexpressiveandefficientOS/NICinterface
Loom'sImplementation
RelatedWork
Evaluation
6
Whyisapplyingthepoliciesdifficult?
Inschedulingpackets,youchoosethenumberofqueuesinthe
NICasaninterfacebetweentheOSandtheNIC
Whenmultiplequeuesareutilized,thepacketsareinflexibly
scheduledinaround-robbinway
Youencounterrespectiveproblemsregardlessofyouchoose
singlequeue(SQ)ormultiplequeues(MQ)
7
ProblemsofSingleQueue(SQ)
Cannotachieveline-rate(100Gbps)duetothecontention
Onlyachievearound36-40Gbpsevenwithbatching
HighCPUutilization
(Butitcanenforcethepolicies)
8
ProblemsofMultipleQueues(MQ)
Itcannotenforcepolicies
Everycoreindependentlyenforcespoliciesonlocaltraffic
Highlatencybytheinflexibleroundrobbinschedulingofthe
queuesforlatency-sensitiveapplications
9
Loom,anewNICdesign
HowtocopewiththeinflexibleNIC?
Solution:moveallper-flowschedulingdecisionsoutoftheOSand
intotheNIC!
Loomdesignchoice
Howshouldpoliciesbespecified?
HowdowebuildaNICthatcanenforceLoomʼsnew
abstraction?
10
Outlines
Motivation
Loom'sDesign
.Anewnetworkpolicyabstraction:restrictedDAGs
.Aprogrammablehierarchicalpacketscheduler
.AnewexpressiveandefficientOS/NICinterface
Loom'sImplementation
RelatedWork
Evaluation
11
DAGpolicyabstraction(1)
NetworkpolicyinLoomcanberepresentedwithschedulingand
shapingnodes
Schedulingnodesdeterminetherelativeorderingofdifferent
packets(WeightedFairQueuing,Prioriyscheduling)
Shapingnodesdeterminethetimingofpackets(rate-limits)
LoomadoptDAGpolicyabstractionofthesekindsofnodes,not
treeabstraction
Treeabstractionisnotabletorepresentcomplexhierarchy
12
DAGpolicyabstraction(2)
NotalltheDAGsarepermitted
DAGswhichformsatreeifallshapingnodesareremovedisonly
allowed
Thisensuresthatschedulingnodesclosertotherootdonot
reorderpacketsthatwerealreadyorderedbyachildnode.
13
Outlines
Motivation
Loom'sDesign
.Anewnetworkpolicyabstraction:restricteddirectedacyclic
graphs(DAGs)
.Aprogrammablehierarchicalpacketscheduler
.AnewexpressiveandefficientOS/NICinterface
Loom'sImplementation
RelatedWork
Evaluation
14
APIFOtree
Manyschedulingalgorithmscanbeimplementedbyatreeof
priotiryqueues(push-infirst-outqueues,PIFOs) 1
Differentschedulingalgorithmsareimplementedbychanginghow
therankiscomputed
Strictpriorityschedulingcomputesapriority
WeightFairQueuingcomputesavirtualtimeforthepacket
Rate-limitingcomputesatransmittime
andsoforth...
1 Sivaraman,A.,Subramanian,S.,Alizadeh,M.,Chole,S.,Chuang,S.T.,Agrawal,A.,Balakrishnan,H.,Edsall,T.,Katti,S.,
&McKeown,N.(2016).Programmablepacketschedulingatlinerate.SIGCOMM2016-Proceedingsofthe2016ACM
ConferenceonSpecialInterestGrouponDataCommunication,44–57.
15
Anewschedulinghierarchy
However,priorPIFO-basedschedulercannotenforcethepolicies
representedasDAGs
Itisonlyapplicabletoatreeofnodesofpolicies
Iamgoingtoexplainthereasonusingexamples
16
PriorPIFO-basedscheduler(1)
Therearequeuesinthehierarchy,theRootqueue,andQ1.
Rate-limitingandsomescheduinginthefigureiscompiled
intoQ1. 17
PriorPIFO-basedscheduler(1)
D1isenqueuedtotheQ1Scheduledqueue.
ApointertoQ1isalsoinsertedintotheRootqueue.
18
PriorPIFO-basedscheduler(1)
ThepointerisdequeuedfromtheRoot,andthepacketatthehead
oftheschedulingqueueatQ1willbereturned.
19
PriorPIFO-basedscheduler(2)
Therearequeuesinthehierarchy,theRootqueue,andQ1.
20
PriorPIFO-basedscheduler(2)
D2,whichhasatransmittimeofT_D2asaresultofrate-limitting,
isinsertedbothQ1SchedulingandShapingqueue
Butitisnotenqueuedhigherupthetreeduetotherate-limiting
21
PriorPIFO-basedscheduler(2)
AttheT_D2,T_D2isdequeuedfromShapingqueue,andapointer
toQ1isinsertedintotheRootqueue.
22
PriorPIFO-basedscheduler(2)
ThepointerisdequeuedfromtheRoot,andthepacketatthehead
oftheschedulingqueueatQ1willbereturned.
23
PriorPIFO-basedscheduler(3)
Usingthisexample,Iwillshowproblemsofenforcementof
policies
BothD1andD2willbeinserted
24
PriorPIFO-basedscheduler(3)
D1isenqueuedtotheScheduledqueue.
ApointertoQ1isalsoinsertedintotheRootqueue.
25
PriorPIFO-basedscheduler(3)
AssumeD2israte-limitedandhashigherrankthanD1
D2isinsertedbothQ1SchedulingandShapingqueue
ItisorderedinQ1aheadofD1consideringthehigherrank
26
PriorPIFO-basedscheduler(3)
ApointerisdequeuedfromtheRoot,andinthistime,thepacketat
theheadoftheschedulingqueueatQ1willbereturned.
ThismeansD2,notD1isdequeuedfromQ1
Thisisnotwhatwehaveexpected.Wewanttorate-limitD2,
notD1
27
PriorPIFO-basedscheduler(3)
D1isenqueuedtotheQ1Scheduledqueue.
ApointertoQ1isalsoinsertedintotheRootqueue.
28
PriorPIFO-basedscheduler(3)
ThepointerisdequeuedfromtheRoot,andthepacketatthehead
oftheschedulingqueueatQ1willbereturned.
29
Howtomodifythescheduler?
Usesaglobalshapingqueuetoimplementrate-limitinginsteadof
per-nodeshapingqueues
Alltrafficisfirstonlyplacedinschedulingqueues
Ifapacketisdequeuedbeforeitsshapingtime,itisplacedin
aglobalshapingqueue
Aftershaping,thepacketisplacedbackinschedulingqueues
30
LoomPIFO-basedscheduler
D1isalreadyinsertedintotheQ1SchedulingqueueandRoot
queue.
31
LoomPIFO-basedscheduler
Thedifferenceis
TheT_D2isnotinsertedintoQ1Shapingqueue
ApointertoQ1isinsertedtoRootqueuethoughD2israte-
limited
32
LoomPIFO-basedscheduler
AdequeueoprationtoRootisconducted
ApointerisdequeuedfromtheRoot,andinthistime,thepacketat
theheadoftheschedulingqueueatQ1willbereturned.
Butthedifferenceis
Itisre-enqueuedinaglobalshapingqueueuntilitsreaches
itstransmissiontimeofT_D2. 33
LoomPIFO-basedscheduler
ThepointerisdequeuedfromtheRoot,andthepacketatthehead
oftheschedulingqueueatQ1willbereturned.
34
LoomPIFO-basedscheduler
AttheT_D2,theD2isre-insertedintothehierarcy.
35
Trade-offofthenewscheduling
Itcanincreasethenumberofenqueueanddequeueoperations
performedperpacket
Intheworstcase,apacketinLoommaybeenqueuedand
dequeuedtwotimesfromthePIFOhierarchy.
TheauthorsinsistthatitdoesnotpreventLoomfromoperatingat
line-rate(100Gbps)
ANICwithfrequencyof1GHzof100Gbpscanschedulea
packetevery2ns.
Achievingline-rateonlyneedstosendapacketevery5.12ns
Otheroptimizationsareexplainedaboutthat
36
Outlines
Motivation
Loom'sDesign
.Anewnetworkpolicyabstraction:restricteddirectedacyclic
graphs(DAGs)
.Aprogrammablehierarchicalpacketscheduler
.AnewexpressiveandefficientOS/NICinterface
Loom'sImplementation
RelatedWork
Evaluation
37
OS/NICInterface
MinimizesthetotalnumberofPCIewritesneededbytheOS
TheOSusesbatcheddoorbellstonotifytheNICofnew
segments.
Schedulingandmetadataupdatesarepassedin-linewith
packetdescriptorstoavoidgeneratingadditionalPCIewrites.
Inthispresentation,Idon'texplainmuchaboutit.
38
Outlines
Motivation
Loom'sDesign
.Anewnetworkpolicyabstraction:restricteddirectedacyclic
graphs(DAGs)
.Aprogrammablehierarchicalpacketscheduler
.AnewexpressiveandefficientOS/NICinterface
Loom'sImplementation
RelatedWork
Evaluation
39
Loom'sImplementation
ThepolicyDAGcompiler
ExistingcompilerforschedulingtreeswritteninDomino 1
SoftwareprototypeLoomNIConBerkeleyExtensibleSoftware
Switch(BESS,formallySoftNIC) 2
1 https://github.com/packet-transactions/domino-compiler
2 http://span.cs.berkeley.edu/bess.html
40
Outlines
Motivation
Loom'sDesign
.Anewnetworkpolicyabstraction:restricteddirectedacyclic
graphs(DAGs)
.Aprogrammablehierarchicalpacketscheduler
.AnewexpressiveandefficientOS/NICinterface
Loom'sImplementation
RelatedWork
Evaluation
41
RelatedWork
SENIC 1
ANICdesignthatcannativelysupportratelimiters
Aqueue-per-corewithunbatcheddoorbells
Requiresynchronizationacrossmultiplecores
Eiffel 2
Improvestheefficiencyofsoftwarepacketscheduling
EiffelandLoomarecomplementary.
1 Radhakrishnan,S.,Geng,Y.,Jeyakumar,V.,Kabbani,A.,Porter,G.,&Vahdat,A.(2014).SENIC:ScalableNICforend-
hostratelimiting.Proceedingsofthe11thUSENIXSymposiumonNetworkedSystemsDesignandImplementation,NSDI
2014,120,475–488.
2 Saeed,A.,Zhao,Y.,Dukkipati,N.,Ammar,M.,Zegura,E.,Harras,K.,&Vahdat,A.(2019).Eiffel:Efficientandflexible
softwarepacketscheduling.Proceedingsofthe16thUSENIXSymposiumonNetworkedSystemsDesignand
Implementation,NSDI2019,17–31.
42
Outlines
Motivation
Loom'sDesign
.Anewnetworkpolicyabstraction:restricteddirectedacyclic
graphs(DAGs)
.Aprogrammablehierarchicalpacketscheduler
.AnewexpressiveandefficientOS/NICinterface
Loom'sImplementation
RelatedWork
Evaluation
43
Evaluation
SendingdatabetweentwoserversonCloudLab 1
Twodifferentexperimentconfigurations
IntelX52010GbENICs
40GbpsMellanoxConnectX-3NICs
ThesoftwareNICusesonecoreforpackettransmissionandone
coreforreception
1 https://www.cloudlab.us/
44
Policyenforcement(1)
Therearefiveactivetenants
Every2sanewtenantstartsorstops
45
Policyenforcement(2)
SQisnotabletofullydrivethe40Gbpsline-rateandMQisnot
ableenforcethispolicy
Loomachievebothline-rateandenforcementofpolicy
46
Policyenforcement(3)
90thpercentilelatencyofT0with10GbENICs
Loomachieveslowlatencyevenwhenthenumberofflows
increases
47
OS/NICinterface
ComparethenumberofPCIewritespersecondgeneratedby
Loom(batcheddoorbell)andSENIC(unbatcheddoorbells).
Inthebestcasescenario(leftfigure),batcheddoorbellswasable
toreducethenumberofwritesbyupto19.2%.
Intheworstcasesenario(rightfigure),thenumberofwritesof
barcheddoorbellbecameupto2.3%.
Existingapproachwillhavedifficultydrivingincreasingline-rates
48
DCapplications
Evaluateisolationwhenusinghierarchicalpolicy
TenantsT1(memcachedandSpark)andT2(Spark)areallocated
equalsharesofbandwidth
Withloom,twoSparkapplicationsachievealmostthesame
bandwidth,andthelatencyofmemcachedisnotimpactedby
Spark
49
Conclusion
ProposeLoom,anewNICdesignthatmovesallper-flow
schedulingdecisionsoutoftheOSandintotheNIC
Loomcontributes
(1)AnewpolicyDAGabstraction
(2)Anewflexibleprogrammableschedulinghierarchy
(3)AnewexpressiveandefficientOS/NICinterface
LoomcancompletelyoffloadallpacketschedulingtotheNICwith
lowCPUoverheadwhilestilldrivingincreasingline-rates
(100Gbps)
Itleadstoreductionsinlatency,increasesinthroughput,and
improvementsinfairnessforcompetingtenantsandapplications
50
Mythroughts
IlikethispaperbecauseitclearlyexplainsthemotivationasNSDIpapersusually
do
ThetechnicalcontributionisextendingthePIFOincrementallyandapplyingthe
optimizationtocancelthetrade-offs
Butthereseembemoretrade-offsthatauthorsdonotmention
TheyinsistthatLoom'sdesigncanachieve100Gbpsbutisitreallythecase?
Thereisnoevaluationof100GbpsNIC,only10Gbpsand40Gbps.They
justgiveexplanationusingplentifulsentencesfor100Gbps.
Partlyduetothelackofmyspeciality,Icannotunderstandwhytheycannot
doexperimentsfor100Gbps
HowcanBESSbypasstheoriginalNICqueuesbecauseitisjustthesoftware-
basedimplementation?Theremustbesomewaystodirectlyconnectthewire.
51

More Related Content

What's hot

The Evolution of Internet-Scale Event Notification Services
The Evolution of Internet-Scale Event Notification ServicesThe Evolution of Internet-Scale Event Notification Services
The Evolution of Internet-Scale Event Notification ServicesHiroshi Ono
 
Preparing to program Aurora at Exascale - Early experiences and future direct...
Preparing to program Aurora at Exascale - Early experiences and future direct...Preparing to program Aurora at Exascale - Early experiences and future direct...
Preparing to program Aurora at Exascale - Early experiences and future direct...inside-BigData.com
 
Real time machine learning proposers day v3
Real time machine learning proposers day v3Real time machine learning proposers day v3
Real time machine learning proposers day v3mustafa sarac
 
Internet Of Things: Vision, Prerequisites and OpenSpime
Internet Of Things: Vision, Prerequisites and OpenSpimeInternet Of Things: Vision, Prerequisites and OpenSpime
Internet Of Things: Vision, Prerequisites and OpenSpimeRoberto Ostinelli
 
IRATI Experimentation, US-EU FIRE Workshop
IRATI Experimentation, US-EU FIRE WorkshopIRATI Experimentation, US-EU FIRE Workshop
IRATI Experimentation, US-EU FIRE WorkshopEleni Trouva
 
5G Cellular D2D RDMA Clusters
5G Cellular D2D RDMA Clusters5G Cellular D2D RDMA Clusters
5G Cellular D2D RDMA ClustersYitzhak Bar-Geva
 
Id. 03 tcp ip model
Id. 03 tcp ip modelId. 03 tcp ip model
Id. 03 tcp ip modelRawa KirKuKi
 
A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...
A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...
A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...Rashid Mijumbi
 
Enea Linux and LWRT FTF China 2012
Enea Linux and LWRT FTF China 2012Enea Linux and LWRT FTF China 2012
Enea Linux and LWRT FTF China 2012EneaSoftware
 
DUNE on current and next generation HPC Platforms
DUNE on current and next generation HPC PlatformsDUNE on current and next generation HPC Platforms
DUNE on current and next generation HPC PlatformsMarkus Blatt
 
RT2014_TCPLA_Nara_27052014-V1
RT2014_TCPLA_Nara_27052014-V1RT2014_TCPLA_Nara_27052014-V1
RT2014_TCPLA_Nara_27052014-V1Andrea PETRUCCI
 
Introduction to Network Function Virtualization (NFV)
Introduction to Network Function Virtualization (NFV)Introduction to Network Function Virtualization (NFV)
Introduction to Network Function Virtualization (NFV)rjain51
 

What's hot (14)

The Evolution of Internet-Scale Event Notification Services
The Evolution of Internet-Scale Event Notification ServicesThe Evolution of Internet-Scale Event Notification Services
The Evolution of Internet-Scale Event Notification Services
 
Preparing to program Aurora at Exascale - Early experiences and future direct...
Preparing to program Aurora at Exascale - Early experiences and future direct...Preparing to program Aurora at Exascale - Early experiences and future direct...
Preparing to program Aurora at Exascale - Early experiences and future direct...
 
Real time machine learning proposers day v3
Real time machine learning proposers day v3Real time machine learning proposers day v3
Real time machine learning proposers day v3
 
Internet Of Things: Vision, Prerequisites and OpenSpime
Internet Of Things: Vision, Prerequisites and OpenSpimeInternet Of Things: Vision, Prerequisites and OpenSpime
Internet Of Things: Vision, Prerequisites and OpenSpime
 
IRATI Experimentation, US-EU FIRE Workshop
IRATI Experimentation, US-EU FIRE WorkshopIRATI Experimentation, US-EU FIRE Workshop
IRATI Experimentation, US-EU FIRE Workshop
 
5G Cellular D2D RDMA Clusters
5G Cellular D2D RDMA Clusters5G Cellular D2D RDMA Clusters
5G Cellular D2D RDMA Clusters
 
Id. 03 tcp ip model
Id. 03 tcp ip modelId. 03 tcp ip model
Id. 03 tcp ip model
 
A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...
A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...
A Connectionist Approach to Dynamic Resource Management for Virtualised Netwo...
 
Enea Linux and LWRT FTF China 2012
Enea Linux and LWRT FTF China 2012Enea Linux and LWRT FTF China 2012
Enea Linux and LWRT FTF China 2012
 
DUNE on current and next generation HPC Platforms
DUNE on current and next generation HPC PlatformsDUNE on current and next generation HPC Platforms
DUNE on current and next generation HPC Platforms
 
RT2014_TCPLA_Nara_27052014-V1
RT2014_TCPLA_Nara_27052014-V1RT2014_TCPLA_Nara_27052014-V1
RT2014_TCPLA_Nara_27052014-V1
 
Openstack meetup NFV
Openstack meetup NFV Openstack meetup NFV
Openstack meetup NFV
 
Introduction to Network Function Virtualization (NFV)
Introduction to Network Function Virtualization (NFV)Introduction to Network Function Virtualization (NFV)
Introduction to Network Function Virtualization (NFV)
 
LPC4300_two_cores
LPC4300_two_coresLPC4300_two_cores
LPC4300_two_cores
 

Similar to Loom: flexible and efficient NIC packet scheduling

The-evolution-of-the-private-cloud
The-evolution-of-the-private-cloudThe-evolution-of-the-private-cloud
The-evolution-of-the-private-cloudGeorge Gilbert
 
Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...
Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...
Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...Toshihiko Yamakami
 
Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...
Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...
Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...Juniper Networks
 
Enabling 3d Microelectronics Platforms Mcms
Enabling 3d Microelectronics Platforms  McmsEnabling 3d Microelectronics Platforms  Mcms
Enabling 3d Microelectronics Platforms McmsIonela
 
Cloud deep-dive0212
Cloud deep-dive0212Cloud deep-dive0212
Cloud deep-dive0212Accenture
 
Future Cloud Infrastructure
Future Cloud InfrastructureFuture Cloud Infrastructure
Future Cloud Infrastructureexponential-inc
 
The Abstracted Network for Industrial Internet
The Abstracted Network for Industrial InternetThe Abstracted Network for Industrial Internet
The Abstracted Network for Industrial InternetMeshDynamics
 
Ericsson Technology Review: The future of cloud computing: Highly distributed...
Ericsson Technology Review: The future of cloud computing: Highly distributed...Ericsson Technology Review: The future of cloud computing: Highly distributed...
Ericsson Technology Review: The future of cloud computing: Highly distributed...Ericsson
 
Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...
Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...
Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...Hidetsugu Sugiyama
 
Software defined networking players
Software defined networking playersSoftware defined networking players
Software defined networking playersAmeer Sameer
 
Juniper Jumpstarts Innovation: Open Sources SDN Controller
Juniper Jumpstarts Innovation: Open Sources SDN ControllerJuniper Jumpstarts Innovation: Open Sources SDN Controller
Juniper Jumpstarts Innovation: Open Sources SDN ControllerJuniper Networks
 
Microsoft Telecommunications Industry News | July 2020
Microsoft Telecommunications Industry News | July 2020Microsoft Telecommunications Industry News | July 2020
Microsoft Telecommunications Industry News | July 2020Rick Lievano
 
small-dumb-cheap-and-copious-the-future-of-the-internet-of-things
small-dumb-cheap-and-copious-the-future-of-the-internet-of-thingssmall-dumb-cheap-and-copious-the-future-of-the-internet-of-things
small-dumb-cheap-and-copious-the-future-of-the-internet-of-thingsMeshDynamics
 
Emerging Computing Architectures
Emerging Computing ArchitecturesEmerging Computing Architectures
Emerging Computing ArchitecturesDaniel Holmberg
 
The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...
The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...
The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...NRB
 
Arrow IoT Quadro Wifi Kit
Arrow IoT Quadro Wifi KitArrow IoT Quadro Wifi Kit
Arrow IoT Quadro Wifi KitAspenCore
 
Virtualizing network services
Virtualizing network servicesVirtualizing network services
Virtualizing network servicesBootcamp SCL
 
International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)inventionjournals
 

Similar to Loom: flexible and efficient NIC packet scheduling (20)

Defining telco cloud at du
Defining telco cloud at du Defining telco cloud at du
Defining telco cloud at du
 
The-evolution-of-the-private-cloud
The-evolution-of-the-private-cloudThe-evolution-of-the-private-cloud
The-evolution-of-the-private-cloud
 
Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...
Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...
Device Stand-by Management Scheme of IoT: A Framework for Dealing with Real-w...
 
Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...
Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...
Juniper Announces Availability of Its Contrail SDN Solution; Showcases Custom...
 
Enabling 3d Microelectronics Platforms Mcms
Enabling 3d Microelectronics Platforms  McmsEnabling 3d Microelectronics Platforms  Mcms
Enabling 3d Microelectronics Platforms Mcms
 
Cloud deep-dive0212
Cloud deep-dive0212Cloud deep-dive0212
Cloud deep-dive0212
 
Future Cloud Infrastructure
Future Cloud InfrastructureFuture Cloud Infrastructure
Future Cloud Infrastructure
 
The Abstracted Network for Industrial Internet
The Abstracted Network for Industrial InternetThe Abstracted Network for Industrial Internet
The Abstracted Network for Industrial Internet
 
Ericsson Technology Review: The future of cloud computing: Highly distributed...
Ericsson Technology Review: The future of cloud computing: Highly distributed...Ericsson Technology Review: The future of cloud computing: Highly distributed...
Ericsson Technology Review: The future of cloud computing: Highly distributed...
 
Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...
Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...
Kubernetes Native Infrastructure and CoreOS Operator Framework for 5G Edge Cl...
 
Software defined networking players
Software defined networking playersSoftware defined networking players
Software defined networking players
 
capitolo 05.pdf
capitolo 05.pdfcapitolo 05.pdf
capitolo 05.pdf
 
Juniper Jumpstarts Innovation: Open Sources SDN Controller
Juniper Jumpstarts Innovation: Open Sources SDN ControllerJuniper Jumpstarts Innovation: Open Sources SDN Controller
Juniper Jumpstarts Innovation: Open Sources SDN Controller
 
Microsoft Telecommunications Industry News | July 2020
Microsoft Telecommunications Industry News | July 2020Microsoft Telecommunications Industry News | July 2020
Microsoft Telecommunications Industry News | July 2020
 
small-dumb-cheap-and-copious-the-future-of-the-internet-of-things
small-dumb-cheap-and-copious-the-future-of-the-internet-of-thingssmall-dumb-cheap-and-copious-the-future-of-the-internet-of-things
small-dumb-cheap-and-copious-the-future-of-the-internet-of-things
 
Emerging Computing Architectures
Emerging Computing ArchitecturesEmerging Computing Architectures
Emerging Computing Architectures
 
The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...
The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...
The NRB Group mainframe day 2021 - IBM Z-Strategy & Roadmap - Adam John Sturg...
 
Arrow IoT Quadro Wifi Kit
Arrow IoT Quadro Wifi KitArrow IoT Quadro Wifi Kit
Arrow IoT Quadro Wifi Kit
 
Virtualizing network services
Virtualizing network servicesVirtualizing network services
Virtualizing network services
 
International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)
 

More from TakuyaFukuoka2

業務で ISUCON することになった話.pdf
業務で ISUCON することになった話.pdf業務で ISUCON することになった話.pdf
業務で ISUCON することになった話.pdfTakuyaFukuoka2
 
Page Fault Support for Network Controllers
Page Fault Support for Network ControllersPage Fault Support for Network Controllers
Page Fault Support for Network ControllersTakuyaFukuoka2
 
Using RDMA Efficiently for Key-Value Services
Using RDMA Efficiently for Key-Value ServicesUsing RDMA Efficiently for Key-Value Services
Using RDMA Efficiently for Key-Value ServicesTakuyaFukuoka2
 
Hyperbolic Caching: Flexible Caching for Web Applications
Hyperbolic Caching: Flexible Caching for Web ApplicationsHyperbolic Caching: Flexible Caching for Web Applications
Hyperbolic Caching: Flexible Caching for Web ApplicationsTakuyaFukuoka2
 
Latency-Tolerant Software Distributed Shared Memory
Latency-Tolerant Software Distributed Shared MemoryLatency-Tolerant Software Distributed Shared Memory
Latency-Tolerant Software Distributed Shared MemoryTakuyaFukuoka2
 
Tardis: Time Traveling Coherence Algorithm for Distributed Shared Memory
Tardis: Time Traveling Coherence Algorithm for Distributed Shared MemoryTardis: Time Traveling Coherence Algorithm for Distributed Shared Memory
Tardis: Time Traveling Coherence Algorithm for Distributed Shared MemoryTakuyaFukuoka2
 
Hybrid MPI: Efficient Message Passing for Multi-core Systems
Hybrid MPI: Efficient Message Passing for Multi-core Systems Hybrid MPI: Efficient Message Passing for Multi-core Systems
Hybrid MPI: Efficient Message Passing for Multi-core Systems TakuyaFukuoka2
 
Overlapping Communication and Computation by Using a Hybrid MPI/SMPSs Approach
Overlapping Communication and Computation by Using a Hybrid MPI/SMPSs ApproachOverlapping Communication and Computation by Using a Hybrid MPI/SMPSs Approach
Overlapping Communication and Computation by Using a Hybrid MPI/SMPSs ApproachTakuyaFukuoka2
 

More from TakuyaFukuoka2 (8)

業務で ISUCON することになった話.pdf
業務で ISUCON することになった話.pdf業務で ISUCON することになった話.pdf
業務で ISUCON することになった話.pdf
 
Page Fault Support for Network Controllers
Page Fault Support for Network ControllersPage Fault Support for Network Controllers
Page Fault Support for Network Controllers
 
Using RDMA Efficiently for Key-Value Services
Using RDMA Efficiently for Key-Value ServicesUsing RDMA Efficiently for Key-Value Services
Using RDMA Efficiently for Key-Value Services
 
Hyperbolic Caching: Flexible Caching for Web Applications
Hyperbolic Caching: Flexible Caching for Web ApplicationsHyperbolic Caching: Flexible Caching for Web Applications
Hyperbolic Caching: Flexible Caching for Web Applications
 
Latency-Tolerant Software Distributed Shared Memory
Latency-Tolerant Software Distributed Shared MemoryLatency-Tolerant Software Distributed Shared Memory
Latency-Tolerant Software Distributed Shared Memory
 
Tardis: Time Traveling Coherence Algorithm for Distributed Shared Memory
Tardis: Time Traveling Coherence Algorithm for Distributed Shared MemoryTardis: Time Traveling Coherence Algorithm for Distributed Shared Memory
Tardis: Time Traveling Coherence Algorithm for Distributed Shared Memory
 
Hybrid MPI: Efficient Message Passing for Multi-core Systems
Hybrid MPI: Efficient Message Passing for Multi-core Systems Hybrid MPI: Efficient Message Passing for Multi-core Systems
Hybrid MPI: Efficient Message Passing for Multi-core Systems
 
Overlapping Communication and Computation by Using a Hybrid MPI/SMPSs Approach
Overlapping Communication and Computation by Using a Hybrid MPI/SMPSs ApproachOverlapping Communication and Computation by Using a Hybrid MPI/SMPSs Approach
Overlapping Communication and Computation by Using a Hybrid MPI/SMPSs Approach
 

Recently uploaded

Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...
Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...
Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...Matt Ray
 
Building Real-Time Data Pipelines: Stream & Batch Processing workshop Slide
Building Real-Time Data Pipelines: Stream & Batch Processing workshop SlideBuilding Real-Time Data Pipelines: Stream & Batch Processing workshop Slide
Building Real-Time Data Pipelines: Stream & Batch Processing workshop SlideChristina Lin
 
Implementing Zero Trust strategy with Azure
Implementing Zero Trust strategy with AzureImplementing Zero Trust strategy with Azure
Implementing Zero Trust strategy with AzureDinusha Kumarasiri
 
Recruitment Management Software Benefits (Infographic)
Recruitment Management Software Benefits (Infographic)Recruitment Management Software Benefits (Infographic)
Recruitment Management Software Benefits (Infographic)Hr365.us smith
 
英国UN学位证,北安普顿大学毕业证书1:1制作
英国UN学位证,北安普顿大学毕业证书1:1制作英国UN学位证,北安普顿大学毕业证书1:1制作
英国UN学位证,北安普顿大学毕业证书1:1制作qr0udbr0
 
What is Advanced Excel and what are some best practices for designing and cre...
What is Advanced Excel and what are some best practices for designing and cre...What is Advanced Excel and what are some best practices for designing and cre...
What is Advanced Excel and what are some best practices for designing and cre...Technogeeks
 
How to Track Employee Performance A Comprehensive Guide.pdf
How to Track Employee Performance A Comprehensive Guide.pdfHow to Track Employee Performance A Comprehensive Guide.pdf
How to Track Employee Performance A Comprehensive Guide.pdfLivetecs LLC
 
Automate your Kamailio Test Calls - Kamailio World 2024
Automate your Kamailio Test Calls - Kamailio World 2024Automate your Kamailio Test Calls - Kamailio World 2024
Automate your Kamailio Test Calls - Kamailio World 2024Andreas Granig
 
Unveiling the Future: Sylius 2.0 New Features
Unveiling the Future: Sylius 2.0 New FeaturesUnveiling the Future: Sylius 2.0 New Features
Unveiling the Future: Sylius 2.0 New FeaturesŁukasz Chruściel
 
EY_Graph Database Powered Sustainability
EY_Graph Database Powered SustainabilityEY_Graph Database Powered Sustainability
EY_Graph Database Powered SustainabilityNeo4j
 
Buds n Tech IT Solutions: Top-Notch Web Services in Noida
Buds n Tech IT Solutions: Top-Notch Web Services in NoidaBuds n Tech IT Solutions: Top-Notch Web Services in Noida
Buds n Tech IT Solutions: Top-Notch Web Services in Noidabntitsolutionsrishis
 
SuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte Germany
SuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte GermanySuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte Germany
SuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte GermanyChristoph Pohl
 
Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...
Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...
Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...OnePlan Solutions
 
Alluxio Monthly Webinar | Cloud-Native Model Training on Distributed Data
Alluxio Monthly Webinar | Cloud-Native Model Training on Distributed DataAlluxio Monthly Webinar | Cloud-Native Model Training on Distributed Data
Alluxio Monthly Webinar | Cloud-Native Model Training on Distributed DataAlluxio, Inc.
 
Cloud Management Software Platforms: OpenStack
Cloud Management Software Platforms: OpenStackCloud Management Software Platforms: OpenStack
Cloud Management Software Platforms: OpenStackVICTOR MAESTRE RAMIREZ
 
Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...
Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...
Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...soniya singh
 
ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...
ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...
ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...Christina Lin
 
GOING AOT WITH GRAALVM – DEVOXX GREECE.pdf
GOING AOT WITH GRAALVM – DEVOXX GREECE.pdfGOING AOT WITH GRAALVM – DEVOXX GREECE.pdf
GOING AOT WITH GRAALVM – DEVOXX GREECE.pdfAlina Yurenko
 
Balasore Best It Company|| Top 10 IT Company || Balasore Software company Odisha
Balasore Best It Company|| Top 10 IT Company || Balasore Software company OdishaBalasore Best It Company|| Top 10 IT Company || Balasore Software company Odisha
Balasore Best It Company|| Top 10 IT Company || Balasore Software company Odishasmiwainfosol
 

Recently uploaded (20)

Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...
Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...
Open Source Summit NA 2024: Open Source Cloud Costs - OpenCost's Impact on En...
 
Building Real-Time Data Pipelines: Stream & Batch Processing workshop Slide
Building Real-Time Data Pipelines: Stream & Batch Processing workshop SlideBuilding Real-Time Data Pipelines: Stream & Batch Processing workshop Slide
Building Real-Time Data Pipelines: Stream & Batch Processing workshop Slide
 
Implementing Zero Trust strategy with Azure
Implementing Zero Trust strategy with AzureImplementing Zero Trust strategy with Azure
Implementing Zero Trust strategy with Azure
 
Recruitment Management Software Benefits (Infographic)
Recruitment Management Software Benefits (Infographic)Recruitment Management Software Benefits (Infographic)
Recruitment Management Software Benefits (Infographic)
 
英国UN学位证,北安普顿大学毕业证书1:1制作
英国UN学位证,北安普顿大学毕业证书1:1制作英国UN学位证,北安普顿大学毕业证书1:1制作
英国UN学位证,北安普顿大学毕业证书1:1制作
 
What is Advanced Excel and what are some best practices for designing and cre...
What is Advanced Excel and what are some best practices for designing and cre...What is Advanced Excel and what are some best practices for designing and cre...
What is Advanced Excel and what are some best practices for designing and cre...
 
How to Track Employee Performance A Comprehensive Guide.pdf
How to Track Employee Performance A Comprehensive Guide.pdfHow to Track Employee Performance A Comprehensive Guide.pdf
How to Track Employee Performance A Comprehensive Guide.pdf
 
Automate your Kamailio Test Calls - Kamailio World 2024
Automate your Kamailio Test Calls - Kamailio World 2024Automate your Kamailio Test Calls - Kamailio World 2024
Automate your Kamailio Test Calls - Kamailio World 2024
 
Unveiling the Future: Sylius 2.0 New Features
Unveiling the Future: Sylius 2.0 New FeaturesUnveiling the Future: Sylius 2.0 New Features
Unveiling the Future: Sylius 2.0 New Features
 
EY_Graph Database Powered Sustainability
EY_Graph Database Powered SustainabilityEY_Graph Database Powered Sustainability
EY_Graph Database Powered Sustainability
 
Buds n Tech IT Solutions: Top-Notch Web Services in Noida
Buds n Tech IT Solutions: Top-Notch Web Services in NoidaBuds n Tech IT Solutions: Top-Notch Web Services in Noida
Buds n Tech IT Solutions: Top-Notch Web Services in Noida
 
SuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte Germany
SuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte GermanySuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte Germany
SuccessFactors 1H 2024 Release - Sneak-Peek by Deloitte Germany
 
Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...
Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...
Maximizing Efficiency and Profitability with OnePlan’s Professional Service A...
 
Alluxio Monthly Webinar | Cloud-Native Model Training on Distributed Data
Alluxio Monthly Webinar | Cloud-Native Model Training on Distributed DataAlluxio Monthly Webinar | Cloud-Native Model Training on Distributed Data
Alluxio Monthly Webinar | Cloud-Native Model Training on Distributed Data
 
Cloud Management Software Platforms: OpenStack
Cloud Management Software Platforms: OpenStackCloud Management Software Platforms: OpenStack
Cloud Management Software Platforms: OpenStack
 
Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...
Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...
Russian Call Girls in Karol Bagh Aasnvi ➡️ 8264348440 💋📞 Independent Escort S...
 
Hot Sexy call girls in Patel Nagar🔝 9953056974 🔝 escort Service
Hot Sexy call girls in Patel Nagar🔝 9953056974 🔝 escort ServiceHot Sexy call girls in Patel Nagar🔝 9953056974 🔝 escort Service
Hot Sexy call girls in Patel Nagar🔝 9953056974 🔝 escort Service
 
ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...
ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...
ODSC - Batch to Stream workshop - integration of Apache Spark, Cassandra, Pos...
 
GOING AOT WITH GRAALVM – DEVOXX GREECE.pdf
GOING AOT WITH GRAALVM – DEVOXX GREECE.pdfGOING AOT WITH GRAALVM – DEVOXX GREECE.pdf
GOING AOT WITH GRAALVM – DEVOXX GREECE.pdf
 
Balasore Best It Company|| Top 10 IT Company || Balasore Software company Odisha
Balasore Best It Company|| Top 10 IT Company || Balasore Software company OdishaBalasore Best It Company|| Top 10 IT Company || Balasore Software company Odisha
Balasore Best It Company|| Top 10 IT Company || Balasore Software company Odisha
 

Loom: flexible and efficient NIC packet scheduling