Analysis Of Server Logs To Determine The Right Timing For Server Maintenance Using A Single Node And Multi-Node Hadoop Environment
Ferhat Oflezer1*, Turgay Tugay Bilgin2
1Bozok Üniversitesi, Boğazlıyan Meslek Yüksekokulu, Bilgisayar Teknolojileri Bölümü , Yozgat, Turkey
2Bilgisayar Mühendisliği, Bursa Teknik Üniversitesi , Bursa, Turkey
* Corresponding author: ferhat.oflezer@bozok.edu.tr
Presented at the Ist International Symposium on Innovative Approaches in Scientific Studies (ISAS 2018), Kemer-Antalya, Turkey, Apr 11, 2018
SETSCI Conference Proceedings, 2018, 2, Page (s): 220-220 , https://doi.org/
Published Date: 23 June 2018 |
1047 
11
Abstract
The log files that are continuously recorded are one of the most important information generated by web
servers.Hadoop is an open source application platform that works on large amounts of data using hardware clusters. Web server
logs are large-volume plain text files that are meaningfully structured. The servers use customized operating systems for resource
sharing. It is extremely important that, some precautions like security updates are formerly planned by monitoring the load status
of the servers over time. Therefore, it is necessary to determine when the server workload is minimal. In the Hadoop environment,
the Hadoop file system is used to store the log files and the results are appreciated using the map and reduce functions. In this
study, the Hadoop MapReduce programming model was used with two mode for analyzing weblogs. These modes are single
node environment mode and distributed environment mode. By using our approach, we have determined how many connections
made to the server during the day and we have tried to find the best time for maintenance.
Keywords - Hadoop; Mapreduce; File System;Bigdata;Data Processing
References