• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.127-148
• /
• 2020

The data center is a physical environment facility for accommodating computer systems and related components, and is an essential foundation technology for next-generation core industries such as big data, smart factories, wearables, and smart homes. In particular, with the growth of cloud computing, the proportional expansion of the data center infrastructure is inevitable. Monitoring the health of these data center facilities is a way to maintain and manage the system and prevent failure. If a failure occurs in some elements of the facility, it may affect not only the relevant equipment but also other connected equipment, and may cause enormous damage. In particular, IT facilities are irregular due to interdependence and it is difficult to know the cause. In the previous study predicting failure in data center, failure was predicted by looking at a single server as a single state without assuming that the devices were mixed. Therefore, in this study, data center failures were classified into failures occurring inside the server (Outage A) and failures occurring outside the server (Outage B), and focused on analyzing complex failures occurring within the server. Server external failures include power, cooling, user errors, etc. Since such failures can be prevented in the early stages of data center facility construction, various solutions are being developed. On the other hand, the cause of the failure occurring in the server is difficult to determine, and adequate prevention has not yet been achieved. In particular, this is the reason why server failures do not occur singularly, cause other server failures, or receive something that causes failures from other servers. In other words, while the existing studies assumed that it was a single server that did not affect the servers and analyzed the failure, in this study, the failure occurred on the assumption that it had an effect between servers. In order to define the complex failure situation in the data center, failure history data for each equipment existing in the data center was used. There are four major failures considered in this study: Network Node Down, Server Down, Windows Activation Services Down, and Database Management System Service Down. The failures that occur for each device are sorted in chronological order, and when a failure occurs in a specific equipment, if a failure occurs in a specific equipment within 5 minutes from the time of occurrence, it is defined that the failure occurs simultaneously. After configuring the sequence for the devices that have failed at the same time, 5 devices that frequently occur simultaneously within the configured sequence were selected, and the case where the selected devices failed at the same time was confirmed through visualization. Since the server resource information collected for failure analysis is in units of time series and has flow, we used Long Short-term Memory (LSTM), a deep learning algorithm that can predict the next state through the previous state. In addition, unlike a single server, the Hierarchical Attention Network deep learning model structure was used in consideration of the fact that the level of multiple failures for each server is different. This algorithm is a method of increasing the prediction accuracy by giving weight to the server as the impact on the failure increases. The study began with defining the type of failure and selecting the analysis target. In the first experiment, the same collected data was assumed as a single server state and a multiple server state, and compared and analyzed. The second experiment improved the prediction accuracy in the case of a complex server by optimizing each server threshold. In the first experiment, which assumed each of a single server and multiple servers, in the case of a single server, it was predicted that three of the five servers did not have a failure even though the actual failure occurred. However, assuming multiple servers, all five servers were predicted to have failed. As a result of the experiment, the hypothesis that there is an effect between servers is proven. As a result of this study, it was confirmed that the prediction performance was superior when the multiple servers were assumed than when the single server was assumed. In particular, applying the Hierarchical Attention Network algorithm, assuming that the effects of each server will be different, played a role in improving the analysis effect. In addition, by applying a different threshold for each server, the prediction accuracy could be improved. This study showed that failures that are difficult to determine the cause can be predicted through historical data, and a model that can predict failures occurring in servers in data centers is presented. It is expected that the occurrence of disability can be prevented in advance using the results of this study.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.10 no.2
• /
• pp.97-114
• /
• 1977

The fresh water prawn, Macrobrachium rosenbergi(de Man) is a very common species in Indopacific region, which inhaits both fresh and brackish water in low land area, most of rivers and especially aboundant in the lower reaches which are influenced by sea water. It is one of the largest and commercial species of genus Macrobrachium, which is commonly larger than $18\~21cm$ in body length, from the basis of eye-stalked to the distal of telson. As a part of the researches in order to investigate the possibilities on transplantation and propagation of this species, this work dealt with the problems on the effects of chlorinities upon zoeal larvae and post-larvae 1). metamorphosis rate and optimum chlorinity for metamorphosis to post-larve, 2). tolerance and comparative survival rate on various chlorinties, from fresh water to sea water $(19.38\%_{\circ}\;Cl)$, which reared for six days upon each stage of zoeal larvae, 3). accomodation rate on chlonities which reared for twelve days after transmigration into variant chlorinities of the range from $3.68\%_{\circ}$ Cl to $1.53\%_{\circ}$ Cl in the way of rearing of the range from $3.82\%_{\circ}$ Cl to $11.05\%_{\circ}$ upon each stage of zoea, 4). tolerance on both of fresh and sea water upon zoeal larva and post-larva under the condition of $28^{\circ}C{\pm}1$ in temperature and feeding on Artenia salina nauplii, 5). relationship between various chlorinities and grwth of post-larvae under the condition of $28^{\circ}C$ in tmperature and feeding on meat of clam. Thus these investigations were performed in order to grope for a comfortable method on seedmass production. Up to the present, the study on the effects of chlorinity upon earlier zoeal larvae and post-larvae of Macrobrachium species has been scarcely performed by workers with the exception of Lewis(1961) and Ling (1962,, 1967), even so their works were not so detailed. On the other hand, larvae of several species of this genus were reared at the water which mixed sea water so as to carry out complete metamorphosis to post-larva by workers in order to investigate on earlier 1 arval and earlier post-larval development, such as Macrobrachium lamerrei (Rajyalakshmi, 1961), M. rosenbergi and M. nipponense (Uno and Kwoa, 1969; Kwon and Uno, 1969), M. acanthurs (Choudhury, 1970; Dobkin, 1971), M. carcinus(Choudhury, 1970), M. formosense(Shokita, 1970), M. olfersii (Duggei et al., 1975), M. novaehallandiae (Greenwood et al., 1976), M. japonicum (Kwon, 1974) and M. lar (Shokita, personal communication), and there fore it is regarded that chlorinity is, generally, one of absolute factors to rear zoeal larvae of brackish species of Macrobrachium genus. Synthetic results on this work is summarized as the follwings: 1) Zoeal larvae required different chlorinities to grow according to each stage, and generally, it is regarded that optimum range of living and growing is from $7.63\%_{\circ}Cl\to\;7.63\%_{\circ}Cl$, and while differences of metamorphsis rate, from first zoea to post-larva, is rarely found in this range, and however it occurs apparently in both of situation at $7.63\%_{\circ}Cl$ below and $16.63\%_{\circ}Cl$ above and moreover, metamorphosis rate is delayed somewhat in case of lower chlorinity as compared with high chlorinity in these situations. 2) Accomodation in each chlorinity on the range, from fresh water to sea water, is different according to larval stages and while the best of it is, generally, on the range from $14.24\%_{\circ}Cl$ to $8.28\%_{\circ}Cl$ and favorite chlorinity of zoea have a tendency to remove from high chlorinity to lower chlorinity in order to advance larval age throughout all zoeal stages, setting a conversional stage for eighta zoea stage. 3) Optimum chlorinity of living and growth upon postlarvae is on the range of $4.25\%_{\circ}Cl$ below, and in proportion as approach to fresh water, growth rate is increased. 4) Post-large are able to live better in fresh water in comparison with zoeal larvae, which are only able to live within fifteen hours, and by contraries, post-larvae are merely able to live for one day as compared with ?미 larvar, which are able to live for six days more in sea water $19.38\%_{\circ}Cl\;above$. 5) Also, in case of transmigration into higher and lower chlorinities in the way of rearing in the initial chlorinities $ 3.82\%_{\circ}Cl,\;7.14%_{\circ}Cl\;and\;11.05%_{\circ}Cl$, accoodation rate is a follow: accomodation capacity in ease of removing into higher chlorinities from lower chlorinities is increased in proportion as earlier stages, setting a conversional stage for eighth zoea stage, and by contraries, in case of advanced stages from eighth zoea it is incraesed in proportion as approach to post-larva stage in the case of transmigration into lower chlorinity from higher chlorinity. On the other hand, it is interesting that in case of reciprocal transmigration between two different chlorinitiess, each survival rate is different, and in this case, also, its accomodation in each zoea stage has a tendency to vary according to larval stages as described above, setting a conversional stage for eighth zoea stage. 6) It is likely that expension of radish pigments on body surface is directly proportional to chlorinity during the period of zoea rearing, and therefore it seems like all body surfacts of zoea larvae be radish coloured in case of higher chlorinity. 7) By the differences that each zoeal larvae, postlarvae, juvaniles and adult prawn are required different chlorinity for inhabiting in each, it is regarded that this species migrats from up steam to near the estuary of the river which the prawns inhabits commonly in natural field for spawning and growth migration. 8) It had better maintainning chlorinities according to zoeal stage for a comfortable method on seed-mass production that earlier larva stages than eighth zoea are maintained on the range from $8\%_{\circ}Cl\;to\;12\%_{\circ}Cl$ to rear, and later larva stages than eighth zoea, by contraries, are gradually regula ted-to love chlorininity of the range from $7\%_{\circ}Cl\;to\;4\%_{\circ}Cl$ according to advance for post-larva stage.