Effective Management of ReRAM-based Hybrid SSD for Multiple Node HDFS

Size: px
Start display at page:

Download "Effective Management of ReRAM-based Hybrid SSD for Multiple Node HDFS"

Transcription

1 Iteratioal Joural of Networked ad Distributed Computig, Vol. 3, No. 3 (August 25), Effective Maagemet of ReRAM-based Hybrid SSD for Multiple HDFS Nayoug Park College of Iformatio ad Commuicatio Egieerig Sugkyukwa Uiversity Suwo, Korea parky42@skku.edu Byugju Lee College of Iformatio ad Commuicatio Egieerig Sugkyukwa Uiversity Suwo, Korea byugju@skku.edu Kyug Tae Kim College of Iformatio ad Commuicatio Egieerig Sugkyukwa Uiversity Suwo, Korea kyugtaekim76@gmail.com Hee Yog You College of Iformatio ad Commuicatio Egieerig Sugkyukwa Uiversity Suwo, Korea you747@skku.edu Abstract Recetly, the research of Hybrid ReRAM/MLC NAND SSD is rapidly expadig ito the storage areas. Most existig researches of Hybrid SSD are based o a sigle storage, while the maagemet of multiple odes like HDFS is still immature. I this paper a ew efficiet cold data evictio scheme is proposed which is based o ode cogestio probability. Computer simulatio reveals that the proposed scheme sigificatly reduces replicatio ad recovery time i compariso to the existig replicatio schemes. Keywords: HDFS, hybrid SSD, ode cogestio probability, replicatio, cold data evictio. Itroductio The demad o efficiet storage for cloud computig has bee growig drastically year after year. Rather tha relyig o traditioal cetralized storage arrays, the storage system for cloud computig cosolidates a large umber of distributed commodity computers ito a sigle storage pool. It provides large capacity ad high performace storage service eve i ureliable ad dyamic etworkig eviromet at low cost. I buildig the cloud storage system, icreasig umber of idustries ad research istitutios rely o the Hadoop Distributed File System (HDFS) []. HDFS provides reliable storage ad high throughput access to the data. It is suitable for the applicatios maipulatig large data sets, typically the oes employig a replicatio maagemet scheme for data-itesive computig. HDFS has bee widely used as a commo storage appliace for cloud computig. Oe of the cospicuous treds i recet storage system is the emergecy of SSD. SSD-based storage is becomig a promisig techology for ext-geeratio storage due to a umber of reasos icludig low access 67

2 latecy, low power cosumptio, higher resistace to shocks, light weight, ad icreasig focus o edurace. Due to the iheret merits of solid state device, SSD ca provide better I/O performace compared with the traditioal HDD. Recetly, a large umber of Iteret service providers have started to replace HDD i their data ceters with SSD [2]. I additio, the fallig cost of NAND flash-based SSD is drivig its expasio ito the market previously reserved for HDD. Fig. shows the architecture of SSD. HostIterface DRAM SSD Cotroller Flash Traslatio Layer Wear Levelig Garbage Collectio Nad Flash Nad Flash Nad Flash Nad Flash Fig.. The block diagram of a traditioal SSD architecture. With the mometum leveraged by both persoal computig ad eterprise system, SSD has bee recogized as a viable choice to build high-performace storage. I order to satisfy the strict requiremets o the future storage system of higher speed, reliability ad eergy-efficiecy, SSD is preferred. Various methods cotributig to the advacemet of the SSD techology have thus bee proposed. Recetly, a hybrid ReRAM/MLC NAND SSD employig the cold data evictio (CDE) algorithm was proposed [3]. It dyamically evicts cold pages from ReRAM to MLC NAND to store hot data. Whe the free space of ReRAM drops below a threshold (for istace, 2% of the ReRAM capacity), evictio is triggered. With this, the page-level migratio is dyamically hadled which is trasparet to the file system. Most existig schemes cosider the evictio based o oly the state of each ode separately. Note, however, that cold ad hot data are dispersed i the odes of Hadoop cluster system. As a result, the performace of the storage system caot be maximized with this approach. Furthermore, it causes uecessary waitig time to the users. If several odes i the HDFS trigger the evictio process simultaeously while may jobs arrive, the users eed to wait for the completio of the evictio process. I this paper, thus, we propose a ew scheme of cold data evictio based o the ode cogestio probability for HDFS. The proposed scheme aalyzes the ode cogestio probability with which the system ca decide if the ode is available to execute the evictio process or ot. I additio, a ew page search approach is proposed, which is importat for high performace HDFS service. Through comprehesive computer simulatio, it is cofirmed that the proposed scheme sigificatly decreases the executio time ad recovery time of replicatio i the HDFS eviromet i compariso to the radom evictio approach, the scheme employed i HDFS, ad the CDE algorithm. The rest of the paper is orgaized as follows. I Sectio 2 brief explaatio of data replicatio with HDFS ad hybrid SSD are provided. Sectio 3 presets the proposed cold data evictio scheme, ad Sectio 4 verifies its performace by computer simulatio. Fially, Sectio 5 cocludes the paper ad suggests the future course of study. 2. Related Work I this sectio the previous work relevat to data replicatio with HDFS ad Hybrid SSD are discussed. 2.. Data Replicatio with HDFS I order to tolerate failure i cloud computig system, various data replicatio techiques have bee proposed. I the cloud computig eviromet, data resources are geographically scattered, ad thus etworkig delay has bee a major obstacle i rapid data access. Numerous studies have bee udertake to replicate data i several data storages that are physically distributed ad as a result reduce the amout of log-distace data trasmissios over the etwork. Fig. 2 describes the structure of data replicatio with HDFS. The data replicatio strategies ca be categorized by the types, uits, ad criteria of replicatio [4]. I terms of replicatio type, there are two types: static ad dyamic. The former is ieffective for large-scale cloud data service because it statically maages data replicatio. It is icapable of quickly respodig to various etwork coditios ad chages i the data access patter. For this reaso, the studies o dyamic data replicatio have bee actively coducted [5, 6]. [7] suggested replicatio strategies reducig the etwork badwidth ad access delay. They also compared the performace of data access patters 68

3 categorized by time ad spatial locality. By cosiderig the etwork capacity ad file access patter, [8] proposed a file replicatio algorithm improvig the performace of basic replicatio method. Similarly, [6] proposed the Latest Access Largest Weight (LALW) method, which uses data access history i dyamically determiig the replicatio policy by applyig greater weight to more recet access. [9] proposed a dyamic optimal replicatio strategy (DORS) which evaluates the value of a file based o the access history, size, ad the etwork coditio i decidig the target files of replicatio. Network Cliets Some researches proposed the strategies for data replicatio focusig o cost-savig. [] proposed a service that applies the market ecoomy model to miimize the replicatio ad data access cost o a data grid. [] suggested a data replicatio strategy based o a cost-estimate model that cosiders both the cost of data access ad the performace of replicatio. I order to esure efficiet resource maagemet i a upredictable data grid eviromet, [2] developed a model that dyamically selects a resource maagemet strategy that respods to a particular workload type based o the performace history. This model demostrated that a real-time workload type is a importat factor i resource maagemet ad the selectio of data replicatio strategy. Similarly, [3] proposed the File Reuio based o Data Replicatio Strategy for Data Grids (FIRE) scheme, which refers to the file access history of earby storage to determie data replicatio, aimig to supply high quality service i the cloud computig eviromet. A data replicatio strategy is dyamically selected to provide optimal data service, while the least Locality based Data Replicatio Strategy (LDRS) scheme tured out to outperform the LRU Meta Data Cotrol Data Blocks Data s Fig. 2. The structure of data replicatio with HDFS. Name scheme for sequetial access patter havig spatial locality. [4] suggested a ovel cost-effective dyamic data replicatio strategy amed CIR for cloud data ceters which applies a icremetal replicatio approach to miimize the umber of replicas while meetig the requiremet of reliability ad cost-effectiveess. Their approach ca substatially reduce the data storage cost, especially whe the data are stored oly for a short duratio or have a lower reliability requiremet. Noetheless, their approach is based o oly the reliability parameters ad pricig model of Amazo S3 which makes it usuitable for Google cluster of a much higher failure rate tha Amazo S3 storage uits. Moreover, they did ot cosider the issue of the tradeoffs betwee cost ad performace. I [5], a dyamic distributed cloud data replicatio algorithm (CDRM) was proposed to capture the relatioship betwee availability ad the umber of replica. It maitais the miimum replica for the give availability requiremet. The replica placemet is based o the capacity ad blockig probability of data odes. Some researches [6] preset six differet replicatio strategies for three differet access patters: No Replicatio or Cachig, Best Cliet, Cascadig Replicatio, Plai Cachig, Cachig plus Cascadig Replicatio, ad Fast Spread. The mai aim of these strategies is reductio i access latecy ad badwidth cosumptio. [7] proposed a cetralized data replicatio algorithm (CDRA) to reduce the total file access time with the cosideratio of limited storage space of Grid sites. Based o the cetralized algorithm, they also desiged a distributed cachig algorithm wherei the Grid sites react close to the Grid status ad make itelliget cachig decisios, which ca be easily adopted i a distributed eviromet such as Data Grids. Their approach ca reduce the aggregated access delay to data files by at least half of that reduced by the optimal replicatio solutio. The limitatio of the algorithm is the cosideratio of oly the access cost. [8] studied a replicatio algorithm based o a costestimatio model, drive by the estimatio of the data access gais ad the replica s creatio ad maiteace cost. It allows the grid odes to automatically replicate data whe eeded i their Data Grid simulator, GridNet. [9] proposed a dyamic hybrid protocol (DHP) which effectively combies the grid ad tree structure so that 69

4 the overall topology ca be flexibly adjusted usig three cofiguratio parameters; tree height, umber of descedats ad grid depth. It ca easily detect read/write coflict ad write/write collisio for cosistecy maiteace. Some of the existig strategies optimize the umber of replicas, while others optimize the placemet of replicas. Others optimize how ofte replicas should be updated [2]. The shortcomig of them is that they oly cosider a restricted set of parameters affectig the replicatio decisio. Furthermore, they oly focus o the improvemet of the system performace without addressig the eergy efficiecy issue i data ceters. [2] desiged a evolutioary way to decide the optimal replicatio strategy. I that work, they optimize storage latecy ad reliability without cosiderig the total eergy cost of data ceter ad the issue of load balacig Hybrid SSD The hybrid SCM/MLC NAND flash SSD (hybrid SSD) is a promisig solutio i boostig the performace of the SSD-based storage while maitaiig the cost. Fig. 3 describes the etire structure of the system implemeted with SSD. SCM operates as both cache ad storage, ot simply as a cache/buffer for the NAND flash memory or merely as storage because it is both fast ad o-volatile. O the host side, the applicatio layer exists above the operatig system, uder which the block device layer resides, which i tur is above the iterface of the host ad storage. The hybrid SSD icludes SSD cotroller, SCM chip array, ad NAND flash memory chip array. Here, MLC NAND flash is preferred to sigle-level cell (SLC) NAND flash to lower the overall cost of the hybrid SSD. As the brai of the storage system, SSD cotroller rus complicated algorithms hadlig the characteristics of erase-before-write ad limited edurace of NAND flash. Withi the cotroller, the data maagemet module determies whether the target data are stored i SCM or i NAND flash memory based o the operatio o the data ad the status of memory. The address traslatio module maages the logical-to-physical address mappig to provide a logical block iterface to the SSD. The wear levelig module guaratees eve wear amog the storage cells to maximize the system logevity. Additioally, the garbage collectio module reclaims free space i the NAND flash whe the NAND flash blocks are almost full. Through the cotroller fuctios, both the performace ad lifetime of SSD ca be ehaced. Fially, the error code correctio (ECC) module detects ad corrects the errors iside the SSD. The role of ECC is becomig more critical as the size scales up, causig degraded reliability. With icreased umber of memory chips, the overhead o the chip area of SSD grows due to the icreased bus area. It is assumed that SCM has much higher reliability tha NAND flash memory. Therefore, oly a pair of simple BCH ECC ecoder/decoder is required for SCM but dozes of low-desity parity-check (LDPC) ECC ecoder/decoders are required for NAND flash. As a result, the area overhead for SSD cotroller due to the icreasig umber of SCM chips is small. I additio, the area for cotrol logic of SCM is egligible. The proposed replicatio scheme for ReRAM-based storage is proposed ext. Hybrid SSD SSD cotroller SCM cotrol Address traslatio ECC NAND flash I/F SCM chips Wear levelig Applicatio layer Operatig system Block device layer SATA,SAS,PCI_e... Data maagemet Chael File system Address traslatio Garbage collectio NAND flash NAND cotrol ECC NAND flash I/F Chael N Wear levelig NAND flash Fig. 3. The overall structure of the whole computer system with SSD. 7

5 3. The Proposed Scheme This sectio proposes a scheme evictig cold data for the HDFS based o the hybrid SSD. It cosists of three parts; evictio of cold data, estimatio of ode cogestio probability used i the cold data evictio, ad page search for fidig cold/hot pages. 3.. System Model Algorithm : Process of cold data evictio The proposed scheme targets the developmet of a replicatio maagemet approach for HDFS system. The HDFS has umerous similarities with the proprietary distributed file system, Google file system. It cosists of a sigle ame ode ad a set of data odes. The ame ode ad data odes are deployed withi a umber of racks as show i Fig. 4, each of which has a associated rack umber. Cotrary to the existig HDFS, i the proposed scheme the ame ode maily maages the amespace of the file system ad the locatio of data pages (the mappig of data pages to data odes). A file is split ito oe or more data pages which are dispersed i the data odes. I Hadoop, the applicatios are executed i data odes. Whe a applicatio eeds a data page, it acts as a HDFS cliet sedig a page read (write) request to the ame ode. The ame ode fids the requested data ode to process the request. Each data ode also periodically seds a heartbeat message to the ame ode to otify its soudess. I HDFS, the umber of replicas is set to three, which is also adopted i this paper. By usig NRU (Not Recetly Used) table, the proposed scheme decides if a data page is hot or cold. Rack Rack 2 Name The overall process of the proposed scheme is give i Fig. 5. The proposed replicatio scheme is triggered if the free space of ReRAM of etire odes drops below a threshold (for istace, 2% of the total ReRAM capacity), the free space of ReRAM of a sigle ode is less tha %, ad the ode cogestio probability (CP) is lower tha the threshold, Tcp. Rack 3 Data Fig. 4. The structure of HDFS. Rack 4 Rack 5 Switch Rack : 2: 3: 4: 5: 6: 7: 8: 9: : : 2: 3: 4: 5: New write data if Total free space i ReRAM < 2% Go to Lie 6 if Sigle ode free space i ReRAM < % if the ode cogestio probability > CPTHRESHOLD if Flag_NRU == (Hit NRU table?) if R<RRM Add page to cold data evictio list else Go to Lie 5 if NRM == RMEXECUTION Executio of cold data evictio else Go to Lie 5 else Search page i ReRAM Fig. 5. The overall process of the proposed scheme. To judge whether a page is hot or cold, the NRU table is referred. Because the write performace is critical to the NAND-based storage system, the proposed algorithm was desiged to optimize the operatio of SSD write request rather tha read request. Here the page utilizatio, U, is compared with the utilizatio threshold, Tu, to choose the least used page for replicatio. If it is impossible to select a page i ReRAM whe cold data evictio is triggered, RRM is reduced to icrease the umber of cadidate pages i ReRAM used i the ext evictio process. Fig. 6 compares the proposed scheme with the existig schemes. I the MLC NAND-oly SSD i Fig. 6(a), all pages are stored i SSD, regardless hot/cold, fragmeted/ot fragmeted. I the covetioal hybrid SSD (Fig. 6(b)) fragmeted pages are set to ReRAM [22]. Sice fragmeted cold data decreases free space of ReRAM, fragmeted hot data might be stored i MLC NAND. As a result, the SSD performace will be greatly degraded if this situatio gets worse or the ReRAM capacity is relatively small. I order to provide high performace uder itesive fragmeted cold data workloads, a ew cold data evictio scheme is proposed. Fig. 6(c) describes that the portios of cold data i ReRAM are evicted to MLC NAND flash. By 7

6 dyamically evictig the fragmeted cold data pages from ReRAM to MLC NAND, ReRAM ca effectively store hot data. Furthermore, fragmetatio i MLC NAND is miimized by the evictio of mostly fragmeted cold data page. MLC NAND oly Host SSD cotroller MLC NAND Covetioal Hybrid SSD Host SSD cotroller ReRAM MLC NAND Data Hot Cold Frag Not frag Proposed Hybrid SSD Host SSD cotroller ReRAM MLC NAND k CP[] i = ωi ACK[]. i ECN -Echo () k i= where ACK[i].ECN-Echo is the value of the CE bit k i ACK[i], ad w i are ormalized such that ωi =. i= The value of CP(t) is calculated usig k serial ACKs received util time t, deoted by ACK[i], ACK[i ],,ACK[i k+], respectively. For the sake of simplicity of presetatio, we use ACK[i] istead of ACK[i].ECN-Echo. The weights of the CE bits i differet ACK packets are assiged usig the expoetially weighted movig average method. The ACKs are divided ito several segmets, ad all ACKs i the same segmet are assiged a same weight. This is because a large umber of ACKs icrease the feedback delay, while a sigle ACK caot correctly reflect the cogestio state. To achieve the tradeoff betwee sesitivity ad stabilizatio, the k ACKs are divided ito segmets, ad the ACKs of segmet_x are assiged with a weight, w x. The value of CP(t) ca be calculated as k ( x ) CP[] i = ω ACK[] i (2) x x= k i= k ( x ) Writte to NAND for ReRAM space shortage 3.2. Cogestio Cold data evictio frees ReRAM space for hot data Fig. 6. The compariso of the proposed scheme with the existig oes. The key idea of cogestio probability based o ECN (Explicit Cogestio Notificatio) is to aalyze the distributio of Cogestio Experieced (CE) bits from the ECN feedbacks ad predict the etwork state accordig to the correlatio betwee multiple feedbacks. Oe ECN feedback reflects the etwork state oly at oe earlier time iterval, while a sequece of ECN feedback ca idicate the dyamic chage of the etwork state durig the past several cotiuous time itervals. Therefore, we propose to combie the iformatio cotaied i the CE bits of a sequece of ACK packets together to predict the etwork state. Assume that a seder receivers a ECN feedback sequece cotaiig k ECN feedbacks. We calculate CP(t), the cogestio probability at the curret time t, as We calculate CP(t) with movig weighted average value of samples. Note that the cogestio probability based o ECN tracks the chage of the CP(t) value to estimate the cogestio level. Geerally, the ewer the ACK packet, the more accurate it is i predictig the curret etwork state. Therefore, a larger weight eeds to be assiged to ewer ACK. The weights of ACKs i each segmet are assiged as ωx = * ωx, α (,) (3) α While all the weights satisfy ωx = (4) x= It is clear that the weight of x th segmet, ω, is larger x tha the weight of previous segmet, ωx, sice the value of α is less tha. I highly dyamic etwork, it is very difficult to get accurate traffic state with oly a smaller umber of 72

7 packets. However, usig large umber of packets will icrease the feedback latecy. Like the way that TCP uses three duplicate ACKs to reflect packet loss, we choose four ACKs for oe segmet to track the cogestio state. After receivig three duplicate ACK packets, the seder eters the cogestio avoidace mode. Similar to this, we choose a period of four ACK packets to compose oe segmet reflectig the cogestio state. Thus, we have k = 4 (5) With Eq. (3), Eq. (4) ad Eq. (5), we ca rewrite Eq. (2) as ( ACK[4] ACK[] ) ωi CP = + ( ACK [8] ACK [4]) 4 α ( ACK[4 ] ACK[4 4]) α ACK[] + ACK[4] α ω i = + ACK[8] 2 4 α α ACK[4 ] α α (8) 4x CP[ i] = ω ACK[ i], i [, k] (6) t x= 4 i= 4( x ) Eq. (6) ca be expaded to 3 CP[] i = ω ACK[] i 4 i= 7 ω + ACK[] i α 4 i= 4 4 ω + + ACK[] i α 4 i= 4( ) (7) If CP(t+) > CP(t), the cogestio probability icreases; otherwise, the cogestio probability decreases. Therefore, the chage of CP ca properly reflect the chage of etwork state. Defiig ΔCP = CP(t+) CP(t) CP(t), we have From Eq. (8), it is clear that ΔCP is related to oly ACK[4i], that is to say, the last ACK of each segmet. We ca also see that the value of ΔCP lies i a discrete set cotaiig 2 + elemets. If the CE bit of the last ACK i the last segmet is marked with, the value of ΔCP will be i [, ]. I this case, it ca be cocluded that the cogestio probability is decreasig, which has the same effect of a sigle ECN feedback. O the cotrary, if the CE bit of the latest arrivig ACK is, the value of ΔCP will be i [, ]. I this case, it caot be determied whether the etwork is goig to be cogested or ot. However, ΔCP always idicates that the etwork cogestio may be relieved regardless of the value of the CE bit of the ewly arrivig ACK. The value of CP is calculated by the procedure show i Fig. 7 Algorithm 2 : Procedure of calculatig CP : CP(t +)=CP(t) 2: it ECN-Echo=hdr_flags::access(pkt) ECN-Echo() 3: float ω, ω 2 4: for i= to m do 5: ACK[i ]=ACK[i] 6: ed for 7: ACK[m ]= ECN-Echo 8: u = ACK[m ]+ +ACK[m/2] 9: u 2 = ACK[m/2 ]+ +ACK[] : CP(t)= ω* u+ ω2* u2 : retur CP(t) Fig. 7. The procedure for calculatig the CP Page Search I searchig cold pages, three approaches are available as show i Fig. 8, (a) From the first page, (b) From radom positio, ad (c) From the usearched page. The umbers i the figure deote the search sequece of ReRAM pages, ad the arrow shows the start poit of the search. For the approach of Fig.8 (c), the start poit is the cold page i ReRAM foud i the previous cold data evictio operatio. The page umber is stored i the memory, ad updated whe the cold data evictio process is triggered. I the proposed scheme the approach of Fig.8 (c) is adopted due to lower search overhead compared with the other approaches. With the adopted search approach, the usearched regio is scaed first. Statistically, the chace of the searched regio to cotai cold page after the executio of cold data evictio will be relatively low. 73

8 By checkig the usearched regio first, the overhead of searchig ca be miimized. We ext evaluate the performace of the proposed scheme. Table. MLC NAND/ReRAM Specificatio ad Used Table MLC NAND ReRAM Previously searched pages Usearched pages Read latecy (max.) 85µs/page s/sector Write latecy(typical) Lower page 4µs Upper page 28µs (Set/Reset) s/sector Erase latecy (typical) 85µs/block Access uit Page(6KiB) Sector(52B) NRU 2 etries (a) From the first ReRAM page (b) From a radom positio (c) From the usearched page Evictio list 5 etries Fig. 8. The approaches for the search of cold page. 4. Performace Evaluatio I the simulatio tree structure is assumed as the etwork topology of the HDFS. To simulate the tree etwork topology, the odes are distributed i the racks as follows: there are racks ad each rack is equipped with oe switch. The racks are radomly distributed over a uit square plae, ad a rack occupies a square plae. For ay two racks, there is o itersectio area betwee them. Amog the racks, oe is desigated as the root rack of all other racks i biary tree topology of the height of 7. After formig the racks i the tree topology, 3,5 odes are radomly deployed withi the racks. For ay two odes i the same rack, their locatios are withi the square plae occupied by the rack. Based o the geerated etwork topology, the simulatio is performed with the parameter settig summarized i Table I. I each ode the available replicatio space is represeted as the maximum umber of data block replicas allowed to be stored. It is set by radomly selectig a umber betwee ad 5 I a simulatio ru, some odes are radomly selected which frequetly eed to write data to their disks. Therefore, may odes ca cocurretly issue the replicatio requests. The umber of requestig odes are chaged from 5 to 2,5 to evaluate the effectiveess of the proposed scheme with differet load coditios. For the settigs of the read, write ad erase latecy, the NRU table ad evictio list are used i additio to the parameters show i Table I. I additio, dyamic workloads are also cosidered for the disk queue ad switch lik queue. Wheever oe or more odes cocurretly issue the replicatio requests, block read-write operatios are eeded betwee multiple ode pairs. The umber of cocurret block read-write operatios is radomly set betwee ad 5. Due to cocurret block read-write operatios, a umber of operatios eed to be hadled with the disk queue ad switch lik queue before issuig the replicatio requests. Total replicatio cost (secods) Radom HDFS CDE Proposed scheme Number of requestig odes Fig. 9. The compariso of total replicatio costs. Fig. 9 shows the total replicatio cost for differet umbers of requestig odes. As see from Fig. 9, the total replicatio cost of all the schemes icrease with the umber of requestig odes. Basically, the HDFS replicatio scheme adopts the radom scheme to place the replicas of a data block, but with additioal cosideratio o the possible rack failure. Therefore, the total replicatio cost of the HDFS replicatio scheme is 74

9 similar to that of the radom replicatio scheme. Notice that the proposed scheme cosistetly reduces the replicatio cost compared to the existig schemes. I case of the requestig odes of 5, the improvemet with the proposed scheme is ot sigificat. However, as the umber of requestig odes icreases, the reductio becomes quite substatial. Average excutio time (secods) Fig.. The compariso of average executio times. Fig. shows that the average executio time follows almost the same tred as the replicatio cost of Fig. 9. The proposed scheme outperforms the other schemes, especially with relatively large umber of requestig odes. Average recovery time (secods) Radom HDFS CDE Proposed scheme Radom CDE Number of requestig odes HDFS Proposed scheme Average Cogestio Probabilities (%) Fig.. Average recovery time with differet cogestio probabilities. The average recovery time is show i Fig. for differet cogestio probabilities ad, requestig odes. As ca be see from the figure that the proposed scheme requires the smallest average recovery time, which is about oe-fifth of that of the HDFS. Ad, as the cogestio probability grows, the improvemet gets larger. Worst recovery time (secods) Fig. 2. The compariso of worst recovery times. Fig. 2 shows the recovery times i the worst case. The worst recovery time deotes the largest access time to retrieve a failure-free data replica, which is measured by accumulatig the access time of all replicas of a data block. As show i the figure, the proposed scheme decreases about 7% of the worst recovery time of the HDFS. Compared to the CDE algorithm, it is about 2%. 5. Coclusio Radom HDFS CDE Proposed scheme Number of requestig odes I this paper we have proposed a ovel cold data evictio scheme which cospicuously cosiders the cogestio status of the odes. The proposed scheme aalyzes the ode cogestio probability by which the system ca decide if the odes are available to trigger the cold data evictio process or ot. Also, the proposed scheme classifies the data as hot or cold, ad decides the storage locatio of them usig NRU algorithm. Computer simulatio reveals that the proposed scheme sigificatly decreases the recovery ad executio time i the HDFS eviromet i compariso to the previous schemes. I particular, with relatively high cogestio probability, the proposed scheme substatia excels the existig schemes. As the future course of study, we pla to exted the proposed scheme to effectively decide the umber ad locatio of the replicatios of SSD blocks i the HDFS eviromet. Also, the desig parameters will be determied through comprehesive modelig of the proposed scheme o the target performace measures. There exist may storage odes i a cloud computig system, ad eergy efficiecy is a importat issue. The 75

10 proposed scheme will be ivestigated further to maximize the eergy efficiecy. Ackowledgemets This research was supported by Basic Sciece Research Program through the Natioal Research Foudatio of Korea (NRF) fuded by the Miistry of Educatio, Sciece ad Techology (22RAA24257 ad 23RAA26398), the secod Brai Korea 2 PLUS project, ICT R&D program of MSIP/IITP (3953), ad Samsug Electroics (S-24-7-). Correspodig author: Hee Yog You. Refereces. K. Shvachko, H. Kuag, S. Radia, ad R. Chasler, The Hadoop Distributed File System, i Proceedigs of the IEEE 26th Symposium o Mass Storage Systems ad Techologies (MSST ), pp.-, May C. Metz. Flash drives replace disks at Amazo, Facebook, Dropbox. Jue C. Su, K. Miyaji, K. Johguchi, K. Takeuchi, A high performace ad eergy-efficiet cold data evictio algorithm for 3D-TSV hybrid ReRAM/MLC NAND SSD, IEEE Trasactios o Circuits ad Systems Ι:Regular papers, vol. 6, o. 2, Feb S. Veugopal, R. Buyya, R. Kotagiri, "A Taxoomy of Data Grids for Distributed Data Sharig, Maagemet ad Processig," ACM Computig Surveys, Vol. 38, pp. -53, K. Sashi, A.S. Thaamai, Dyamic Replicatio i a Data Grid Usig a Modified BHR Regio Based Algorithm, Future Geeratio Computer Systems, Vol. 27, No. 2, pp. 22-2, 2 6. R.S. Chag, H.P. Chag, "A Dyamic Data Replicatio Strategy Usig Access-Weights i Data Grids," Supercomputig, Vol. 45, No. 3, pp , K. Ragaatha, I. Foster, "Desig ad Evaluatio of Dyamic Replicatio Strategies for a High-Performace Data Grid," Iteratioal Coferece o Computig i High Eergy ad Nuclear Physics, H. Sato, et al., "Access-Patter ad Badwidth Aware File Replicatio Algorithm i a Grid Eviromet," Iteratioal Coferece o Grid Computig, pp , W. Zhao, et al., A Dyamic Optimal Replicatio Strategy i Data Grid Eviromet, Iteratioal Coferece o Iteret Techology ad Applicatios, pp. -4, 2.. M. Carma, K. Stockiger, "Toward a Ecoomy-based Optimizatio of File Access ad Replicatio o a Data Grid," Iteratioal Symposium o Cluster Computig ad the Grid, pp , 22.. H. Lamehamedi, et al., "Data Replicatio Strategies i Grid Eviromets", Iteratioal Coferece o Algorithms ad Architectures for Parallel Processig, pp , B.D. Lee, J.B. Weissma, Y.K. Nam, "Adaptive Middleware Supportig Scalable Performace for High- Ed Network Service," Network ad Computer Applicatios, Vol. 32, pp , A.R. Abdurrab, T. Xie, "FIRE: A File Reuio Based Data Replicatio Strategy for Data Grids," Iteratioal Coferece o Clusterig Computig ad the Grid, pp , W.H. Li, Y. Yag, D. Yua, A ovel cost-effective dyamic data replicatio strategy for reliability i cloud data cetres, i: IEEE Nith Iteratioal Coferece o Depedable, Autoomic ad Secure Computig, Q. Wei, B. Veeravalli, B. Gog, L. Zeg, D. Feg, CDRM: a cost-effective dyamic replicatio maagemet scheme for cloud storage cluster, i: Proc. 2 IEEE Iteratioal Coferece o Cluster Computig, Heraklio, Crete, Greece,September 2 24, 2, pp K. Ragaatha, I.T. Foster, Idetifyig dyamic replicatio strategies for a high-performace data grid, i: Proc. Secod It l Workshop Grid Computig (GRID), D.T. Nukarapu, B. Tag, L.Q. Wag, S.Y. Lu, Data replicatio i data itesive scietific applicatios with performace guaratee, IEEE Tras. Parallel Distrib. Syst. 22 (8) (2) H. Lamehamedi, Z. Shetu, B. Szymaski, Simulatio of dyamic data replicatio strategies i data grids, i: Proc. 2th Heterogeeous Computig Workshop (HCW23) Nice, Frace, April 23, IEEE Computer Sciece Press, Los Alamitos, CA, S.C. Choi, H.Y. You, Dyamic hybrid replicatio effectively combiig tree ad grid topology, J. Supercomput. 59 (22) M. Tu, T. Tadayo, Z. Xia, E. Lu, A secure ad scalable update protocol for P2P data grids, i: th IEEE High Assurace Systems Egieerig Symposium, Texas, 27, pp O.A.-H. Hassa, L. Ramaswamy, J. Miller, K. Rasheed, E.R. Cafield, Replicatio i overlay etworks: a multiobjective optimizatio approach, i: Collaborative Computig: Networkig, Applicatios ad Worksharig, Lecture Notes of the Istitute for Computer Scieces, Social Iformatics ad Telecommuicatios Egieerig, vol., 29, pp H. Fujii, K. Miyaji, K. Johguchi, K. Higuchi, C. Su, ad K. Takeuchi, performace icrease, 6.9 edurace ehacemet, 93% eergy reductio of 3D TSVitegrated hybrid ReRAM/MLC NAND SSDs by data fragmetatio suppressio, i Symp. VLSI Circuits Dig.Tech. Papers, Ju. 22, pp

An Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem

An Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem A Improved Shuffled Frog-Leapig Algorithm for Kapsack Problem Zhoufag Li, Ya Zhou, ad Peg Cheg School of Iformatio Sciece ad Egieerig Hea Uiversity of Techology ZhegZhou, Chia lzhf1978@126.com Abstract.

More information

UH-MEM: Utility-Based Hybrid Memory Management. Yang Li, Saugata Ghose, Jongmoo Choi, Jin Sun, Hui Wang, Onur Mutlu

UH-MEM: Utility-Based Hybrid Memory Management. Yang Li, Saugata Ghose, Jongmoo Choi, Jin Sun, Hui Wang, Onur Mutlu UH-MEM: Utility-Based Hybrid Memory Maagemet Yag Li, Saugata Ghose, Jogmoo Choi, Ji Su, Hui Wag, Our Mutlu 1 Executive Summary DRAM faces sigificat techology scalig difficulties Emergig memory techologies

More information

Master Informatics Eng. 2017/18. A.J.Proença. Memory Hierarchy. (most slides are borrowed) AJProença, Advanced Architectures, MiEI, UMinho, 2017/18 1

Master Informatics Eng. 2017/18. A.J.Proença. Memory Hierarchy. (most slides are borrowed) AJProença, Advanced Architectures, MiEI, UMinho, 2017/18 1 Advaced Architectures Master Iformatics Eg. 2017/18 A.J.Proeça Memory Hierarchy (most slides are borrowed) AJProeça, Advaced Architectures, MiEI, UMiho, 2017/18 1 Itroductio Programmers wat ulimited amouts

More information

A QoS Provisioning mechanism of Real-time Wireless USB Transfers for Smart HDTV Multimedia Services

A QoS Provisioning mechanism of Real-time Wireless USB Transfers for Smart HDTV Multimedia Services A QoS Provisioig mechaism of Real-time Wireless USB Trasfers for Smart HDTV Multimedia Services Ji-Woo im 1, yeog Hur 2, Jog-Geu Jeog 3, Dog Hoo Lee 4, Moo Sog Yeu 5, Yeowoo Lee 6 ad Seog Ro Lee 7 1 Istitute

More information

CMSC Computer Architecture Lecture 11: More Caches. Prof. Yanjing Li University of Chicago

CMSC Computer Architecture Lecture 11: More Caches. Prof. Yanjing Li University of Chicago CMSC 22200 Computer Architecture Lecture 11: More Caches Prof. Yajig Li Uiversity of Chicago Lecture Outlie Caches 2 Review Memory hierarchy Cache basics Locality priciples Spatial ad temporal How to access

More information

Quorum Based Data Replication in Grid Environment

Quorum Based Data Replication in Grid Environment Quorum Based Data Replicatio i Grid Eviromet Rohaya Latip, Hamidah Ibrahim, Mohamed Othma, Md Nasir Sulaima, ad Azizol Abdullah Faculty of Computer Sciece ad Iformatio Techology, Uiversiti Putra Malaysia

More information

3D Model Retrieval Method Based on Sample Prediction

3D Model Retrieval Method Based on Sample Prediction 20 Iteratioal Coferece o Computer Commuicatio ad Maagemet Proc.of CSIT vol.5 (20) (20) IACSIT Press, Sigapore 3D Model Retrieval Method Based o Sample Predictio Qigche Zhag, Ya Tag* School of Computer

More information

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5.

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5. Morga Kaufma Publishers 26 February, 208 COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Chapter 5 Virtual Memory Review: The Memory Hierarchy Take advatage of the priciple

More information

Adaptive Resource Allocation for Electric Environmental Pollution through the Control Network

Adaptive Resource Allocation for Electric Environmental Pollution through the Control Network Available olie at www.sciecedirect.com Eergy Procedia 6 (202) 60 64 202 Iteratioal Coferece o Future Eergy, Eviromet, ad Materials Adaptive Resource Allocatio for Electric Evirometal Pollutio through the

More information

CMSC Computer Architecture Lecture 10: Caches. Prof. Yanjing Li University of Chicago

CMSC Computer Architecture Lecture 10: Caches. Prof. Yanjing Li University of Chicago CMSC 22200 Computer Architecture Lecture 10: Caches Prof. Yajig Li Uiversity of Chicago Midterm Recap Overview ad fudametal cocepts ISA Uarch Datapath, cotrol Sigle cycle, multi cycle Pipeliig Basic idea,

More information

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5 Morga Kaufma Publishers 26 February, 28 COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Chapter 5 Set-Associative Cache Architecture Performace Summary Whe CPU performace icreases:

More information

Sectio 4, a prototype project of settig field weight with AHP method is developed ad the experimetal results are aalyzed. Fially, we coclude our work

Sectio 4, a prototype project of settig field weight with AHP method is developed ad the experimetal results are aalyzed. Fially, we coclude our work 200 2d Iteratioal Coferece o Iformatio ad Multimedia Techology (ICIMT 200) IPCSIT vol. 42 (202) (202) IACSIT Press, Sigapore DOI: 0.7763/IPCSIT.202.V42.0 Idex Weight Decisio Based o AHP for Iformatio Retrieval

More information

Analysis of Server Resource Consumption of Meteorological Satellite Application System Based on Contour Curve

Analysis of Server Resource Consumption of Meteorological Satellite Application System Based on Contour Curve Advaces i Computer, Sigals ad Systems (2018) 2: 19-25 Clausius Scietific Press, Caada Aalysis of Server Resource Cosumptio of Meteorological Satellite Applicatio System Based o Cotour Curve Xiagag Zhao

More information

Prevention of Black Hole Attack in Mobile Ad-hoc Networks using MN-ID Broadcasting

Prevention of Black Hole Attack in Mobile Ad-hoc Networks using MN-ID Broadcasting Vol.2, Issue.3, May-Jue 2012 pp-1017-1021 ISSN: 2249-6645 Prevetio of Black Hole Attack i Mobile Ad-hoc Networks usig MN-ID Broadcastig Atoy Devassy 1, K. Jayathi 2 *(PG scholar, ME commuicatio Systems,

More information

Harris Corner Detection Algorithm at Sub-pixel Level and Its Application Yuanfeng Han a, Peijiang Chen b * and Tian Meng c

Harris Corner Detection Algorithm at Sub-pixel Level and Its Application Yuanfeng Han a, Peijiang Chen b * and Tian Meng c Iteratioal Coferece o Computatioal Sciece ad Egieerig (ICCSE 015) Harris Corer Detectio Algorithm at Sub-pixel Level ad Its Applicatio Yuafeg Ha a, Peijiag Che b * ad Tia Meg c School of Automobile, Liyi

More information

Course Site: Copyright 2012, Elsevier Inc. All rights reserved.

Course Site:   Copyright 2012, Elsevier Inc. All rights reserved. Course Site: http://cc.sjtu.edu.c/g2s/site/aca.html 1 Computer Architecture A Quatitative Approach, Fifth Editio Chapter 2 Memory Hierarchy Desig 2 Outlie Memory Hierarchy Cache Desig Basic Cache Optimizatios

More information

Low Complexity H.265/HEVC Coding Unit Size Decision for a Videoconferencing System

Low Complexity H.265/HEVC Coding Unit Size Decision for a Videoconferencing System BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 15, No 6 Special Issue o Logistics, Iformatics ad Service Sciece Sofia 2015 Prit ISSN: 1311-9702; Olie ISSN: 1314-4081 DOI:

More information

Energy-efficient HTTP Adaptive Streaming for High- Quality Video over HetNets

Energy-efficient HTTP Adaptive Streaming for High- Quality Video over HetNets 5 IEEE 6th Iteratioal Symposium o Persoal Idoor ad Mobile Radio Commuicatios - (PIMRC): Services Eergy-efficiet HTTP Adaptive Streamig for High- Quality Video over HetNets Yumi Go Oh Cha Kwo ad Hwaju Sog

More information

Ones Assignment Method for Solving Traveling Salesman Problem

Ones Assignment Method for Solving Traveling Salesman Problem Joural of mathematics ad computer sciece 0 (0), 58-65 Oes Assigmet Method for Solvig Travelig Salesma Problem Hadi Basirzadeh Departmet of Mathematics, Shahid Chamra Uiversity, Ahvaz, Ira Article history:

More information

FAST BIT-REVERSALS ON UNIPROCESSORS AND SHARED-MEMORY MULTIPROCESSORS

FAST BIT-REVERSALS ON UNIPROCESSORS AND SHARED-MEMORY MULTIPROCESSORS SIAM J. SCI. COMPUT. Vol. 22, No. 6, pp. 2113 2134 c 21 Society for Idustrial ad Applied Mathematics FAST BIT-REVERSALS ON UNIPROCESSORS AND SHARED-MEMORY MULTIPROCESSORS ZHAO ZHANG AND XIAODONG ZHANG

More information

Optimization for framework design of new product introduction management system Ma Ying, Wu Hongcui

Optimization for framework design of new product introduction management system Ma Ying, Wu Hongcui 2d Iteratioal Coferece o Electrical, Computer Egieerig ad Electroics (ICECEE 2015) Optimizatio for framework desig of ew product itroductio maagemet system Ma Yig, Wu Hogcui Tiaji Electroic Iformatio Vocatioal

More information

Algorithms for Disk Covering Problems with the Most Points

Algorithms for Disk Covering Problems with the Most Points Algorithms for Disk Coverig Problems with the Most Poits Bi Xiao Departmet of Computig Hog Kog Polytechic Uiversity Hug Hom, Kowloo, Hog Kog csbxiao@comp.polyu.edu.hk Qigfeg Zhuge, Yi He, Zili Shao, Edwi

More information

Elementary Educational Computer

Elementary Educational Computer Chapter 5 Elemetary Educatioal Computer. Geeral structure of the Elemetary Educatioal Computer (EEC) The EEC coforms to the 5 uits structure defied by vo Neuma's model (.) All uits are preseted i a simplified

More information

performance to the performance they can experience when they use the services from a xed location.

performance to the performance they can experience when they use the services from a xed location. I the Proceedigs of The First Aual Iteratioal Coferece o Mobile Computig ad Networkig (MobiCom 9) November -, 99, Berkeley, Califoria USA Performace Compariso of Mobile Support Strategies Rieko Kadobayashi

More information

n Explore virtualization concepts n Become familiar with cloud concepts

n Explore virtualization concepts n Become familiar with cloud concepts Chapter Objectives Explore virtualizatio cocepts Become familiar with cloud cocepts Chapter #15: Architecture ad Desig 2 Hypervisor Virtualizatio ad cloud services are becomig commo eterprise tools to

More information

CMSC Computer Architecture Lecture 12: Virtual Memory. Prof. Yanjing Li University of Chicago

CMSC Computer Architecture Lecture 12: Virtual Memory. Prof. Yanjing Li University of Chicago CMSC 22200 Computer Architecture Lecture 12: Virtual Memory Prof. Yajig Li Uiversity of Chicago A System with Physical Memory Oly Examples: most Cray machies early PCs Memory early all embedded systems

More information

Improvement of the Orthogonal Code Convolution Capabilities Using FPGA Implementation

Improvement of the Orthogonal Code Convolution Capabilities Using FPGA Implementation Improvemet of the Orthogoal Code Covolutio Capabilities Usig FPGA Implemetatio Naima Kaabouch, Member, IEEE, Apara Dhirde, Member, IEEE, Saleh Faruque, Member, IEEE Departmet of Electrical Egieerig, Uiversity

More information

Adaptive and Lazy Segmentation Based Proxy Caching for Streaming Media Delivery

Adaptive and Lazy Segmentation Based Proxy Caching for Streaming Media Delivery Adaptive ad Based Proxy Cachig for Streamig Media Delivery Sogqig Che Dept. of Computer Sciece College of William ad Mary Williamsburg, VA 23187 sqche@cs.wm.edu Susie Wee Mobile ad Media System Lab Hewlett-Packard

More information

1. SWITCHING FUNDAMENTALS

1. SWITCHING FUNDAMENTALS . SWITCING FUNDMENTLS Switchig is the provisio of a o-demad coectio betwee two ed poits. Two distict switchig techiques are employed i commuicatio etwors-- circuit switchig ad pacet switchig. Circuit switchig

More information

Enhancing Cloud Computing Scheduling based on Queuing Models

Enhancing Cloud Computing Scheduling based on Queuing Models Ehacig Cloud Computig Schedulig based o Queuig Models Mohamed Eisa Computer Sciece Departmet, Port Said Uiversity, 42526 Port Said, Egypt E. I. Esedimy Computer Sciece Departmet, Masoura Uiversity, Masoura,

More information

High Availability with Diagonal Replication in 2D Mesh (DR2M) Protocol for Grid Environment

High Availability with Diagonal Replication in 2D Mesh (DR2M) Protocol for Grid Environment High Availability with Diagoal Replicatio i 2D Mesh (DR2M) Protocol for Grid Eviromet Rohaya Latip (Correspodig author) Faculty of Computer Sciece ad Iformatio Techology Uiversiti Putra Malaysia Tel: 60-3-8946-6536

More information

Basic allocator mechanisms The course that gives CMU its Zip! Memory Management II: Dynamic Storage Allocation Mar 6, 2000.

Basic allocator mechanisms The course that gives CMU its Zip! Memory Management II: Dynamic Storage Allocation Mar 6, 2000. 5-23 The course that gives CM its Zip Memory Maagemet II: Dyamic Storage Allocatio Mar 6, 2000 Topics Segregated lists Buddy system Garbage collectio Mark ad Sweep Copyig eferece coutig Basic allocator

More information

Appendix D. Controller Implementation

Appendix D. Controller Implementation COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Appedix D Cotroller Implemetatio Cotroller Implemetatios Combiatioal logic (sigle-cycle); Fiite state machie (multi-cycle, pipelied);

More information

Chapter 1. Introduction to Computers and C++ Programming. Copyright 2015 Pearson Education, Ltd.. All rights reserved.

Chapter 1. Introduction to Computers and C++ Programming. Copyright 2015 Pearson Education, Ltd.. All rights reserved. Chapter 1 Itroductio to Computers ad C++ Programmig Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Overview 1.1 Computer Systems 1.2 Programmig ad Problem Solvig 1.3 Itroductio to C++ 1.4 Testig

More information

Operating System Concepts. Operating System Concepts

Operating System Concepts. Operating System Concepts Chapter 4: Mass-Storage Systems Logical Disk Structure Logical Disk Structure Disk Schedulig Disk Maagemet RAID Structure Disk drives are addressed as large -dimesioal arrays of logical blocks, where the

More information

A New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method

A New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method A ew Morphological 3D Shape Decompositio: Grayscale Iterframe Iterpolatio Method D.. Vizireau Politehica Uiversity Bucharest, Romaia ae@comm.pub.ro R. M. Udrea Politehica Uiversity Bucharest, Romaia mihea@comm.pub.ro

More information

Lecture 28: Data Link Layer

Lecture 28: Data Link Layer Automatic Repeat Request (ARQ) 2. Go ack N ARQ Although the Stop ad Wait ARQ is very simple, you ca easily show that it has very the low efficiecy. The low efficiecy comes from the fact that the trasmittig

More information

MAC Throughput Improvement Using Adaptive Contention Window

MAC Throughput Improvement Using Adaptive Contention Window Joural of Computer ad Commuicatios, 2015, 3, 1 14 Published Olie Jauary 2015 i SciRes. http://www.scirp.org/joural/jcc http://dx.doi.org/10.4236/jcc.2015.31001 MAC Throughput Improvemet Usig Adaptive Cotetio

More information

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe Copyright 2016 Ramez Elmasri ad Shamkat B. Navathe CHAPTER 19 Query Optimizatio Copyright 2016 Ramez Elmasri ad Shamkat B. Navathe Itroductio Query optimizatio Coducted by a query optimizer i a DBMS Goal:

More information

1 Enterprise Modeler

1 Enterprise Modeler 1 Eterprise Modeler Itroductio I BaaERP, a Busiess Cotrol Model ad a Eterprise Structure Model for multi-site cofiguratios are itroduced. Eterprise Structure Model Busiess Cotrol Models Busiess Fuctio

More information

HADOOP: A NEW APPROACH FOR DOCUMENT CLUSTERING

HADOOP: A NEW APPROACH FOR DOCUMENT CLUSTERING Y.K. Patil* Iteratioal Joural of Advaced Research i ISSN: 2278-6244 IT ad Egieerig Impact Factor: 4.54 HADOOP: A NEW APPROACH FOR DOCUMENT CLUSTERING Prof. V.S. Nadedkar** Abstract: Documet clusterig is

More information

AN EFFICIENT GROUP KEY MANAGEMENT USING CODE FOR KEY CALCULATION FOR SIMULTANEOUS JOIN/LEAVE: CKCS

AN EFFICIENT GROUP KEY MANAGEMENT USING CODE FOR KEY CALCULATION FOR SIMULTANEOUS JOIN/LEAVE: CKCS Iteratioal Joural of Computer Networks & Commuicatios (IJCNC) Vol.4, No.4, July 01 AN EFFICIENT GROUP KEY MANAGEMENT USING CODE FOR KEY CALCULATION FOR SIMULTANEOUS JOIN/LEAVE: CKCS Melisa Hajyvahabzadeh

More information

. Written in factored form it is easy to see that the roots are 2, 2, i,

. Written in factored form it is easy to see that the roots are 2, 2, i, CMPS A Itroductio to Programmig Programmig Assigmet 4 I this assigmet you will write a java program that determies the real roots of a polyomial that lie withi a specified rage. Recall that the roots (or

More information

Chapter 4 Threads. Operating Systems: Internals and Design Principles. Ninth Edition By William Stallings

Chapter 4 Threads. Operating Systems: Internals and Design Principles. Ninth Edition By William Stallings Operatig Systems: Iterals ad Desig Priciples Chapter 4 Threads Nith Editio By William Stalligs Processes ad Threads Resource Owership Process icludes a virtual address space to hold the process image The

More information

Mobile terminal 3D image reconstruction program development based on Android Lin Qinhua

Mobile terminal 3D image reconstruction program development based on Android Lin Qinhua Iteratioal Coferece o Automatio, Mechaical Cotrol ad Computatioal Egieerig (AMCCE 05) Mobile termial 3D image recostructio program developmet based o Adroid Li Qihua Sichua Iformatio Techology College

More information

Load balanced Parallel Prime Number Generator with Sieve of Eratosthenes on Cluster Computers *

Load balanced Parallel Prime Number Generator with Sieve of Eratosthenes on Cluster Computers * Load balaced Parallel Prime umber Geerator with Sieve of Eratosthees o luster omputers * Soowook Hwag*, Kyusik hug**, ad Dogseug Kim* *Departmet of Electrical Egieerig Korea Uiversity Seoul, -, Rep. of

More information

Cache-Optimal Methods for Bit-Reversals

Cache-Optimal Methods for Bit-Reversals Proceedigs of the ACM/IEEE Supercomputig Coferece, November 1999, Portlad, Orego, U.S.A. Cache-Optimal Methods for Bit-Reversals Zhao Zhag ad Xiaodog Zhag Departmet of Computer Sciece College of William

More information

Interference Aware Channel Assignment Scheme in Multichannel Wireless Mesh Networks

Interference Aware Channel Assignment Scheme in Multichannel Wireless Mesh Networks Iterferece Aware Chael Assigmet Scheme i Multichael Wireless Mesh Networks Sumyeg Kim Departmet of Computer Software Egieerig Kumoh Natioal Istitute of Techology Gum South Korea Abstract Wireless mesh

More information

CS2410 Computer Architecture. Flynn s Taxonomy

CS2410 Computer Architecture. Flynn s Taxonomy CS2410 Computer Architecture Dept. of Computer Sciece Uiversity of Pittsburgh http://www.cs.pitt.edu/~melhem/courses/2410p/idex.html 1 Fly s Taxoomy SISD Sigle istructio stream Sigle data stream (SIMD)

More information

The University of Adelaide, School of Computer Science 22 November Computer Architecture. A Quantitative Approach, Sixth Edition.

The University of Adelaide, School of Computer Science 22 November Computer Architecture. A Quantitative Approach, Sixth Edition. Computer Architecture A Quatitative Approach, Sixth Editio Chapter 2 Memory Hierarchy Desig 1 Itroductio Programmers wat ulimited amouts of memory with low latecy Fast memory techology is more expesive

More information

S. Mehta and K.S. Kwak. UWB Wireless Communications Research Center, Inha University Incheon, , Korea

S. Mehta and K.S. Kwak. UWB Wireless Communications Research Center, Inha University Incheon, , Korea S. Mehta ad K.S. Kwak UWB Wireless Commuicatios Research Ceter, Iha Uiversity Icheo, 402-75, Korea suryaad.m@gmail.com ABSTRACT I this paper, we propose a hybrid medium access cotrol protocol (H-MAC) for

More information

Multi-Threading. Hyper-, Multi-, and Simultaneous Thread Execution

Multi-Threading. Hyper-, Multi-, and Simultaneous Thread Execution Multi-Threadig Hyper-, Multi-, ad Simultaeous Thread Executio 1 Performace To Date Icreasig processor performace Pipeliig. Brach predictio. Super-scalar executio. Out-of-order executio. Caches. Hyper-Threadig

More information

Evaluation scheme for Tracking in AMI

Evaluation scheme for Tracking in AMI A M I C o m m u i c a t i o A U G M E N T E D M U L T I - P A R T Y I N T E R A C T I O N http://www.amiproject.org/ Evaluatio scheme for Trackig i AMI S. Schreiber a D. Gatica-Perez b AMI WP4 Trackig:

More information

Euclidean Distance Based Feature Selection for Fault Detection Prediction Model in Semiconductor Manufacturing Process

Euclidean Distance Based Feature Selection for Fault Detection Prediction Model in Semiconductor Manufacturing Process Vol.133 (Iformatio Techology ad Computer Sciece 016), pp.85-89 http://dx.doi.org/10.1457/astl.016. Euclidea Distace Based Feature Selectio for Fault Detectio Predictio Model i Semicoductor Maufacturig

More information

A Key Distribution method for Reducing Storage and Supporting High Level Security in the Large-scale WSN

A Key Distribution method for Reducing Storage and Supporting High Level Security in the Large-scale WSN Iteratioal Joural of Digital Cotet Techology ad its Applicatios Vol. 2 No 1, March 2008 A Key Distributio method for Reducig Storage ad Supportig High Level Security i the Large-scale WSN Yoo-Su Jeog *1,

More information

Session Initiated Protocol (SIP) and Message-based Load Balancing (MBLB)

Session Initiated Protocol (SIP) and Message-based Load Balancing (MBLB) F5 White Paper Sessio Iitiated Protocol (SIP) ad Message-based Load Balacig (MBLB) The ability to provide ew ad creative methods of commuicatios has esured a SIP presece i almost every orgaizatio. The

More information

A Study on the Performance of Cholesky-Factorization using MPI

A Study on the Performance of Cholesky-Factorization using MPI A Study o the Performace of Cholesky-Factorizatio usig MPI Ha S. Kim Scott B. Bade Departmet of Computer Sciece ad Egieerig Uiversity of Califoria Sa Diego {hskim, bade}@cs.ucsd.edu Abstract Cholesky-factorizatio

More information

SCI Reflective Memory

SCI Reflective Memory Embedded SCI Solutios SCI Reflective Memory (Experimetal) Atle Vesterkjær Dolphi Itercoect Solutios AS Olaf Helsets vei 6, N-0621 Oslo, Norway Phoe: (47) 23 16 71 42 Fax: (47) 23 16 71 80 Mail: atleve@dolphiics.o

More information

Accuracy Improvement in Camera Calibration

Accuracy Improvement in Camera Calibration Accuracy Improvemet i Camera Calibratio FaJie L Qi Zag ad Reihard Klette CITR, Computer Sciece Departmet The Uiversity of Aucklad Tamaki Campus, Aucklad, New Zealad fli006, qza001@ec.aucklad.ac.z r.klette@aucklad.ac.z

More information

n Learn how resiliency strategies reduce risk n Discover automation strategies to reduce risk

n Learn how resiliency strategies reduce risk n Discover automation strategies to reduce risk Chapter Objectives Lear how resiliecy strategies reduce risk Discover automatio strategies to reduce risk Chapter #16: Architecture ad Desig Resiliecy ad Automatio Strategies 2 Automatio/Scriptig Resiliet

More information

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe

Copyright 2016 Ramez Elmasri and Shamkant B. Navathe Copyright 2016 Ramez Elmasri ad Shamkat B. Navathe CHAPTER 22 Database Recovery Techiques Copyright 2016 Ramez Elmasri ad Shamkat B. Navathe Itroductio Recovery algorithms Recovery cocepts Write-ahead

More information

Software development of components for complex signal analysis on the example of adaptive recursive estimation methods.

Software development of components for complex signal analysis on the example of adaptive recursive estimation methods. Software developmet of compoets for complex sigal aalysis o the example of adaptive recursive estimatio methods. SIMON BOYMANN, RALPH MASCHOTTA, SILKE LEHMANN, DUNJA STEUER Istitute of Biomedical Egieerig

More information

Service Oriented Enterprise Architecture and Service Oriented Enterprise

Service Oriented Enterprise Architecture and Service Oriented Enterprise Approved for Public Release Distributio Ulimited Case Number: 09-2786 The 23 rd Ope Group Eterprise Practitioers Coferece Service Orieted Eterprise ad Service Orieted Eterprise Ya Zhao, PhD Pricipal, MITRE

More information

CSC 220: Computer Organization Unit 11 Basic Computer Organization and Design

CSC 220: Computer Organization Unit 11 Basic Computer Organization and Design College of Computer ad Iformatio Scieces Departmet of Computer Sciece CSC 220: Computer Orgaizatio Uit 11 Basic Computer Orgaizatio ad Desig 1 For the rest of the semester, we ll focus o computer architecture:

More information

Ontology-based Decision Support System with Analytic Hierarchy Process for Tour Package Selection

Ontology-based Decision Support System with Analytic Hierarchy Process for Tour Package Selection 2017 Asia-Pacific Egieerig ad Techology Coferece (APETC 2017) ISBN: 978-1-60595-443-1 Otology-based Decisio Support System with Aalytic Hierarchy Process for Tour Pacage Selectio Tie-We Sug, Chia-Jug Lee,

More information

Partition in mobile adhoc network with mobility-a new approach for effective use of fast IP address autoconfiguration

Partition in mobile adhoc network with mobility-a new approach for effective use of fast IP address autoconfiguration Mukesh Chad et al. / Iteratioal Joural of Egieerig ad Techology (IJET) Partitio i mobile adhoc etwork with mobility-a ew approach for effective use of fast IP address autocofiguratio Mukesh Chad *, HL

More information

arxiv: v2 [cs.ds] 24 Mar 2018

arxiv: v2 [cs.ds] 24 Mar 2018 Similar Elemets ad Metric Labelig o Complete Graphs arxiv:1803.08037v [cs.ds] 4 Mar 018 Pedro F. Felzeszwalb Brow Uiversity Providece, RI, USA pff@brow.edu March 8, 018 We cosider a problem that ivolves

More information

Image Segmentation EEE 508

Image Segmentation EEE 508 Image Segmetatio Objective: to determie (etract) object boudaries. It is a process of partitioig a image ito distict regios by groupig together eighborig piels based o some predefied similarity criterio.

More information

Dynamic Programming and Curve Fitting Based Road Boundary Detection

Dynamic Programming and Curve Fitting Based Road Boundary Detection Dyamic Programmig ad Curve Fittig Based Road Boudary Detectio SHYAM PRASAD ADHIKARI, HYONGSUK KIM, Divisio of Electroics ad Iformatio Egieerig Chobuk Natioal Uiversity 664-4 Ga Deokji-Dog Jeoju-City Jeobuk

More information

Pattern Recognition Systems Lab 1 Least Mean Squares

Pattern Recognition Systems Lab 1 Least Mean Squares Patter Recogitio Systems Lab 1 Least Mea Squares 1. Objectives This laboratory work itroduces the OpeCV-based framework used throughout the course. I this assigmet a lie is fitted to a set of poits usig

More information

Announcements. Reading. Project #4 is on the web. Homework #1. Midterm #2. Chapter 4 ( ) Note policy about project #3 missing components

Announcements. Reading. Project #4 is on the web. Homework #1. Midterm #2. Chapter 4 ( ) Note policy about project #3 missing components Aoucemets Readig Chapter 4 (4.1-4.2) Project #4 is o the web ote policy about project #3 missig compoets Homework #1 Due 11/6/01 Chapter 6: 4, 12, 24, 37 Midterm #2 11/8/01 i class 1 Project #4 otes IPv6Iit,

More information

Politecnico di Milano Advanced Network Technologies Laboratory. Internet of Things. Projects

Politecnico di Milano Advanced Network Technologies Laboratory. Internet of Things. Projects Politecico di Milao Advaced Network Techologies Laboratory Iteret of Thigs Projects 2016-2017 Politecico di Milao Advaced Network Techologies Laboratory Geeral Rules Geeral Rules o Gradig 26/30 are assiged

More information

GTS Scheduling Scheme for Real-Time Communication in IEEE Industrial Wireless Sensor Networks

GTS Scheduling Scheme for Real-Time Communication in IEEE Industrial Wireless Sensor Networks Idia Joural of Sciece ad Techology, Vol 9(7), DOI:.7485/ijst/6/v9i7/87734, February 6 ISSN (Prit) : 974-6846 ISSN (Olie) : 974-5645 GTS Schedulig Scheme for Real-Time Commuicatio i IEEE8.5.4 Idustrial

More information

A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON

A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON Roberto Lopez ad Eugeio Oñate Iteratioal Ceter for Numerical Methods i Egieerig (CIMNE) Edificio C1, Gra Capitá s/, 08034 Barceloa, Spai ABSTRACT I this work

More information

1&1 Next Level Hosting

1&1 Next Level Hosting 1&1 Next Level Hostig Performace Level: Performace that grows with your requiremets Copyright 1&1 Iteret SE 2017 1ad1.com 2 1&1 NEXT LEVEL HOSTING 3 Fast page loadig ad short respose times play importat

More information

Computer Architecture ELEC3441

Computer Architecture ELEC3441 CPU-Memory Bottleeck Computer Architecture ELEC44 CPU Memory Lecture 8 Cache Dr. Hayde Kwok-Hay So Departmet of Electrical ad Electroic Egieerig Performace of high-speed computers is usually limited by

More information

World Scientific Research Journal (WSRJ) ISSN: Research on Fresnel Lens Optical Receiving Antenna in Indoor Visible

World Scientific Research Journal (WSRJ) ISSN: Research on Fresnel Lens Optical Receiving Antenna in Indoor Visible World Scietific Research Joural (WSRJ) ISSN: 2472-3703 www.wsr-j.org Research o Fresel Les Optical Receivig Atea i Idoor Visible Light Commuicatio Zhihua Du College of Electroics Egieerig, Chogqig Uiversity

More information

CIS 121 Data Structures and Algorithms with Java Spring Stacks and Queues Monday, February 12 / Tuesday, February 13

CIS 121 Data Structures and Algorithms with Java Spring Stacks and Queues Monday, February 12 / Tuesday, February 13 CIS Data Structures ad Algorithms with Java Sprig 08 Stacks ad Queues Moday, February / Tuesday, February Learig Goals Durig this lab, you will: Review stacks ad queues. Lear amortized ruig time aalysis

More information

GPUMP: a Multiple-Precision Integer Library for GPUs

GPUMP: a Multiple-Precision Integer Library for GPUs GPUMP: a Multiple-Precisio Iteger Library for GPUs Kaiyog Zhao ad Xiaowe Chu Departmet of Computer Sciece, Hog Kog Baptist Uiversity Hog Kog, P. R. Chia Email: {kyzhao, chxw}@comp.hkbu.edu.hk Abstract

More information

FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS

FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS Prosejit Bose Evagelos Kraakis Pat Mori Yihui Tag School of Computer Sciece, Carleto Uiversity {jit,kraakis,mori,y

More information

ISSN (Print) Research Article. *Corresponding author Nengfa Hu

ISSN (Print) Research Article. *Corresponding author Nengfa Hu Scholars Joural of Egieerig ad Techology (SJET) Sch. J. Eg. Tech., 2016; 4(5):249-253 Scholars Academic ad Scietific Publisher (A Iteratioal Publisher for Academic ad Scietific Resources) www.saspublisher.com

More information

A QoS balancing model for Mobile Ad hoc Networks

A QoS balancing model for Mobile Ad hoc Networks A QoS balacig model for Mobile Ad hoc Networks HENG LUO, DAVID. I. LAURENSON, AI-HUANG GUO, JIANPING CHEN, PEIDE YANG, HUA JIANG Suzhou key laboratory of mobile etworkig ad applied techologies SuZhou Uiversity

More information

EFFECT OF QUERY FORMATION ON WEB SEARCH ENGINE RESULTS

EFFECT OF QUERY FORMATION ON WEB SEARCH ENGINE RESULTS Iteratioal Joural o Natural Laguage Computig (IJNLC) Vol. 2, No., February 203 EFFECT OF QUERY FORMATION ON WEB SEARCH ENGINE RESULTS Raj Kishor Bisht ad Ila Pat Bisht 2 Departmet of Computer Sciece &

More information

Task scenarios Outline. Scenarios in Knowledge Extraction. Proposed Framework for Scenario to Design Diagram Transformation

Task scenarios Outline. Scenarios in Knowledge Extraction. Proposed Framework for Scenario to Design Diagram Transformation 6-0-0 Kowledge Trasformatio from Task Scearios to View-based Desig Diagrams Nima Dezhkam Kamra Sartipi {dezhka, sartipi}@mcmaster.ca Departmet of Computig ad Software McMaster Uiversity CANADA SEKE 08

More information

BASED ON ITERATIVE ERROR-CORRECTION

BASED ON ITERATIVE ERROR-CORRECTION A COHPARISO OF CRYPTAALYTIC PRICIPLES BASED O ITERATIVE ERROR-CORRECTIO Miodrag J. MihaljeviC ad Jova Dj. GoliC Istitute of Applied Mathematics ad Electroics. Belgrade School of Electrical Egieerig. Uiversity

More information

Neuro Fuzzy Model for Human Face Expression Recognition

Neuro Fuzzy Model for Human Face Expression Recognition IOSR Joural of Computer Egieerig (IOSRJCE) ISSN : 2278-0661 Volume 1, Issue 2 (May-Jue 2012), PP 01-06 Neuro Fuzzy Model for Huma Face Expressio Recogitio Mr. Mayur S. Burage 1, Prof. S. V. Dhopte 2 1

More information

Research Article Optimal Configuration of Virtual Links for Avionics Network Systems

Research Article Optimal Configuration of Virtual Links for Avionics Network Systems Iteratioal Joural of Aerospace Egieerig Volume 205, Article ID 329820, 9 pages http://dx.doi.org/0.55/205/329820 Research Article Optimal Cofiguratio of Virtual Liks for Avioics Network Systems Dogha A,

More information

Introduction. Nature-Inspired Computing. Terminology. Problem Types. Constraint Satisfaction Problems - CSP. Free Optimization Problem - FOP

Introduction. Nature-Inspired Computing. Terminology. Problem Types. Constraint Satisfaction Problems - CSP. Free Optimization Problem - FOP Nature-Ispired Computig Hadlig Costraits Dr. Şima Uyar September 2006 Itroductio may practical problems are costraied ot all combiatios of variable values represet valid solutios feasible solutios ifeasible

More information

Pruning and Summarizing the Discovered Time Series Association Rules from Mechanical Sensor Data Qing YANG1,a,*, Shao-Yu WANG1,b, Ting-Ting ZHANG2,c

Pruning and Summarizing the Discovered Time Series Association Rules from Mechanical Sensor Data Qing YANG1,a,*, Shao-Yu WANG1,b, Ting-Ting ZHANG2,c Advaces i Egieerig Research (AER), volume 131 3rd Aual Iteratioal Coferece o Electroics, Electrical Egieerig ad Iformatio Sciece (EEEIS 2017) Pruig ad Summarizig the Discovered Time Series Associatio Rules

More information

Hashing Functions Performance in Packet Classification

Hashing Functions Performance in Packet Classification Hashig Fuctios Performace i Packet Classificatio Mahmood Ahmadi ad Stepha Wog Computer Egieerig Laboratory Faculty of Electrical Egieerig, Mathematics ad Computer Sciece Delft Uiversity of Techology {mahmadi,

More information

Architectural styles for software systems The client-server style

Architectural styles for software systems The client-server style Architectural styles for software systems The cliet-server style Prof. Paolo Ciacarii Software Architecture CdL M Iformatica Uiversità di Bologa Ageda Cliet server style CS two tiers CS three tiers CS

More information

Page 1. Why Care About the Memory Hierarchy? Memory. DRAMs over Time. Virtual Memory!

Page 1. Why Care About the Memory Hierarchy? Memory. DRAMs over Time. Virtual Memory! Why Care About the Memory Hierarchy? Memory Virtual Memory -DRAM Memory Gap (latecy) Reasos: Multi process systems (abstractio & memory protectio) Solutio: Tables (holdig per process traslatios) Fast traslatio

More information

The identification of key quality characteristics based on FAHP

The identification of key quality characteristics based on FAHP Iteratioal Joural of Research i Egieerig ad Sciece (IJRES ISSN (Olie: 2320-9364, ISSN (Prit: 2320-9356 Volume 3 Issue 6 ǁ Jue 2015 ǁ PP.01-07 The idetificatio of ey quality characteristics based o FAHP

More information

CIS 121 Data Structures and Algorithms with Java Spring Stacks, Queues, and Heaps Monday, February 18 / Tuesday, February 19

CIS 121 Data Structures and Algorithms with Java Spring Stacks, Queues, and Heaps Monday, February 18 / Tuesday, February 19 CIS Data Structures ad Algorithms with Java Sprig 09 Stacks, Queues, ad Heaps Moday, February 8 / Tuesday, February 9 Stacks ad Queues Recall the stack ad queue ADTs (abstract data types from lecture.

More information

Analysis Metrics. Intro to Algorithm Analysis. Slides. 12. Alg Analysis. 12. Alg Analysis

Analysis Metrics. Intro to Algorithm Analysis. Slides. 12. Alg Analysis. 12. Alg Analysis Itro to Algorithm Aalysis Aalysis Metrics Slides. Table of Cotets. Aalysis Metrics 3. Exact Aalysis Rules 4. Simple Summatio 5. Summatio Formulas 6. Order of Magitude 7. Big-O otatio 8. Big-O Theorems

More information

Energy Efficinet Cluster Head Selection and Data Convening in Wireless Sensor Networks

Energy Efficinet Cluster Head Selection and Data Convening in Wireless Sensor Networks Idia Joural of Sciece ad Techology, Vol 9(15), DOI: 1.17485/ijst/216/v9i15/77749, April 216 ISSN (Prit) : 974-6846 ISSN (Olie) : 974-5645 Eergy Efficiet Cluster Head Selectio ad Data Coveig i Wireless

More information

Structuring Redundancy for Fault Tolerance. CSE 598D: Fault Tolerant Software

Structuring Redundancy for Fault Tolerance. CSE 598D: Fault Tolerant Software Structurig Redudacy for Fault Tolerace CSE 598D: Fault Tolerat Software What do we wat to achieve? Versios Damage Assessmet Versio 1 Error Detectio Iputs Versio 2 Voter Outputs State Restoratio Cotiued

More information

On Nonblocking Folded-Clos Networks in Computer Communication Environments

On Nonblocking Folded-Clos Networks in Computer Communication Environments O Noblockig Folded-Clos Networks i Computer Commuicatio Eviromets Xi Yua Departmet of Computer Sciece, Florida State Uiversity, Tallahassee, FL 3306 xyua@cs.fsu.edu Abstract Folded-Clos etworks, also referred

More information

Enhancing Efficiency of Software Fault Tolerance Techniques in Satellite Motion System

Enhancing Efficiency of Software Fault Tolerance Techniques in Satellite Motion System Joural of Iformatio Systems ad Telecommuicatio, Vol. 2, No. 3, July-September 2014 173 Ehacig Efficiecy of Software Fault Tolerace Techiques i Satellite Motio System Hoda Baki Departmet of Electrical ad

More information

Switching Hardware. Spring 2018 CS 438 Staff, University of Illinois 1

Switching Hardware. Spring 2018 CS 438 Staff, University of Illinois 1 Switchig Hardware Sprig 208 CS 438 Staff, Uiversity of Illiois Where are we? Uderstad Differet ways to move through a etwork (forwardig) Read sigs at each switch (datagram) Follow a kow path (virtual circuit)

More information